Tag Archives: Data Analysis, Statistics and Probability

Panchromatic simulated galaxy observations from the NIHAO project [GA]

Posted on by

http://arxiv.org/abs/2305.10232


We present simulated galaxy spectral energy distributions (SEDs) from the far ultraviolet through the far infrared, created using hydrodynamic simulations and radiative transfer calculations, suitable for the validation of SED modeling techniques. SED modeling is an essential tool for inferring star formation histories from nearby galaxy observations, but is fraught with difficulty due to our incomplete understanding of stellar populations, chemical enrichment processes, and the non-linear, geometry dependent effects of dust on our observations. Our simulated SEDs will allow us to assess the accuracy of these inferences against galaxies with known ground truth. To create the SEDs, we use simulated galaxies from the NIHAO suite and the radiative transfer code SKIRT. We explore different sub-grid post-processing recipes, using color distributions and their dependence on axis ratio of galaxies in the nearby universe to tune and validate them. We find that sub-grid post-processing recipes that mitigate limitations in the temporal and spatial resolution of the simulations are required for producing FUV to FIR photometry that statistically reproduce the colors of galaxies in the nearby universe. With this paper we release resolved photometry and spatially integrated spectra for our sample galaxies, each from a range of different viewing angles. Our simulations predict that there is a large variation in attenuation laws among galaxies, and that from any particular viewing angle that energy balance between dust attenuation and reemission can be violated by up to a factor of 3. These features are likely to affect SED modeling accuracy.

Read this paper on arXiv…

N. Faucher, M. Blanton and A. Macciò
Thu, 18 May 23
19/67

Comments: N/A

How to estimate Fisher matrices from simulations [CL]

Posted on by

http://arxiv.org/abs/2305.08994


The Fisher information matrix is a quantity of fundamental importance for information geometry and asymptotic statistics. In practice, it is widely used to quickly estimate the expected information available in a data set and guide experimental design choices. In many modern applications, it is intractable to analytically compute the Fisher information and Monte Carlo methods are used instead. The standard Monte Carlo method produces estimates of the Fisher information that can be biased when the Monte-Carlo noise is non-negligible. Most problematic is noise in the derivatives as this leads to an overestimation of the available constraining power, given by the inverse Fisher information. In this work we find another simple estimate that is oppositely biased and produces an underestimate of the constraining power. This estimator can either be used to give approximate bounds on the parameter constraints or can be combined with the standard estimator to give improved, approximately unbiased estimates. Both the alternative and the combined estimators are asymptotically unbiased so can be also used as a convergence check of the standard approach. We discuss potential limitations of these estimators and provide methods to assess their reliability. These methods accelerate the convergence of Fisher forecasts, as unbiased estimates can be achieved with fewer Monte Carlo samples, and so can be used to reduce the simulated data set size by several orders of magnitude.

Read this paper on arXiv…

W. Coulton and B. Wandelt
Wed, 17 May 23
29/67

Comments: Supporting code available at this https URL

Partition function approach to non-Gaussian likelihoods: physically motivated convergence criteria for Markov-chains [CEA]

Posted on by

http://arxiv.org/abs/2305.07061


Non-Gaussian distributions in cosmology are commonly evaluated with Monte Carlo Markov-chain methods, as the Fisher-matrix formalism is restricted to the Gaussian case. The Metropolis-Hastings algorithm will provide samples from the posterior distribution after a burn-in period, and the corresponding convergence is usually quantified with the Gelman-Rubin criterion. In this paper, we investigate the convergence of the Metropolis-Hastings algorithm by drawing analogies to statistical Hamiltonian systems in thermal equilibrium for which a canonical partition sum exists. Specifically, we quantify virialisation, equipartition and thermalisation of Hamiltonian Monte Carlo Markov-chains for a toy-model and for the likelihood evaluation for a simple dark energy model constructed from supernova data. We follow the convergence of these criteria to the values expected in thermal equilibrium, in comparison to the Gelman-Rubin criterion. We find that there is a much larger class of physically motivated convergence criteria with clearly defined target values indicating convergence. As a numerical tool, we employ physics-informed neural networks for speeding up the sampling process.

Read this paper on arXiv…

L. Röver, H. Campe, M. Herzog, et. al.
Mon, 15 May 23
24/53

Comments: 12 pages, 6 figures

Weakly-Supervised Anomaly Detection in the Milky Way [GA]

Posted on by

http://arxiv.org/abs/2305.03761


Large-scale astrophysics datasets present an opportunity for new machine learning techniques to identify regions of interest that might otherwise be overlooked by traditional searches. To this end, we use Classification Without Labels (CWoLa), a weakly-supervised anomaly detection method, to identify cold stellar streams within the more than one billion Milky Way stars observed by the Gaia satellite. CWoLa operates without the use of labeled streams or knowledge of astrophysical principles. Instead, we train a classifier to distinguish between mixed samples for which the proportions of signal and background samples are unknown. This computationally lightweight strategy is able to detect both simulated streams and the known stream GD-1 in data. Originally designed for high-energy collider physics, this technique may have broad applicability within astrophysics as well as other domains interested in identifying localized anomalies.

Read this paper on arXiv…

M. Pettee, S. Thanvantri, B. Nachman, et. al.
Tue, 9 May 23
48/88

Comments: N/A

Machine Learning and Structure Formation in Modified Gravity [CEA]

Posted on by

http://arxiv.org/abs/2305.02122


In General Relativity approximations based on the spherical collapse model such as Press–Schechter theory and its extensions are able to predict the number of objects of a certain mass in a given volume. In this paper we use a machine learning algorithm to test whether such approximations hold in screened modified gravity theories. To this end, we train random forest classifiers on data from N-body simulations to study the formation of structures in $\Lambda$CDM as well as screened modified gravity theories, in particular $f(R)$ and nDGP gravity. The models are taught to distinguish structure membership in the final conditions from spherical aggregations of density field behaviour in the initial conditions. We examine the differences between machine learning models that have learned structure formation from each gravity, as well as the model that has learned from $\Lambda$CDM. We also test the generalisability of the $\Lambda$CDM model on data from $f(R)$ and nDGP gravities of varying strengths, and therefore the generalisability of Extended-Press-Schechter spherical collapse to these types of modified gravity.

Read this paper on arXiv…

J. Betts, C. Bruck, C. Arnold, et. al.
Thu, 4 May 23
21/60

Comments: 8 pages, 6 figures

Interpretable Machine Learning for Science with PySR and SymbolicRegression.jl [IMA]

Posted on by

http://arxiv.org/abs/2305.01582


PySR is an open-source library for practical symbolic regression, a type of machine learning which aims to discover human-interpretable symbolic models. PySR was developed to democratize and popularize symbolic regression for the sciences, and is built on a high-performance distributed back-end, a flexible search algorithm, and interfaces with several deep learning packages. PySR’s internal search algorithm is a multi-population evolutionary algorithm, which consists of a unique evolve-simplify-optimize loop, designed for optimization of unknown scalar constants in newly-discovered empirical expressions. PySR’s backend is the extremely optimized Julia library SymbolicRegression.jl, which can be used directly from Julia. It is capable of fusing user-defined operators into SIMD kernels at runtime, performing automatic differentiation, and distributing populations of expressions to thousands of cores across a cluster. In describing this software, we also introduce a new benchmark, “EmpiricalBench,” to quantify the applicability of symbolic regression algorithms in science. This benchmark measures recovery of historical empirical equations from original and synthetic datasets.

Read this paper on arXiv…

M. Cranmer
Wed, 3 May 23
23/67

Comments: 24 pages, 5 figures, 3 tables. Feedback welcome. Paper source found at this https URL ; PySR at this https URL ; SymbolicRegression.jl at this https URL

Growing Pains: Understanding the Impact of Likelihood Uncertainty on Hierarchical Bayesian Inference for Gravitational-Wave Astronomy [IMA]

Posted on by

http://arxiv.org/abs/2304.06138


Observations of gravitational waves emitted by merging compact binaries have provided tantalising hints about stellar astrophysics, cosmology, and fundamental physics. However, the physical parameters describing the systems, (mass, spin, distance) used to extract these inferences about the Universe are subject to large uncertainties. The current method of performing these analyses requires performing many Monte Carlo integrals to marginalise over the uncertainty in the properties of the individual binaries and the survey selection bias. These Monte Carlo integrals are subject to fundamental statistical uncertainties. Previous treatments of this statistical uncertainty has focused on ensuring the precision of the inferred inference is unaffected, however, these works have neglected the question of whether sufficient accuracy can also be achieved. In this work, we provide a practical exploration of the impact of uncertainty in our analyses and provide a suggested framework for verifying that astrophysical inferences made with the gravitational-wave transient catalogue are accurate. Applying our framework to models used by the LIGO-Virgo-Kagra collaboration, we find that Monte Carlo uncertainty in estimating the survey selection bias is the limiting factor in our ability to probe narrow populations model and this will rapidly grow more problematic as the size of the observed population increases.

Read this paper on arXiv…

C. Talbot and J. Golomb
Fri, 14 Apr 23
52/64

Comments: 8 pages, 6 figures

LyAl-Net: A high-efficiency Lyman-$α$ forest simulation with a neural network [CEA]

Posted on by

http://arxiv.org/abs/2303.17939


The inference of cosmological quantities requires accurate and large hydrodynamical cosmological simulations. Unfortunately, their computational time can take millions of CPU hours for a modest coverage in cosmological scales ($\approx (100 {h^{-1}}\,\text{Mpc})^3)$). The possibility to generate large quantities of mock Lyman-$\alpha$ observations opens up the possibility of much better control on covariance matrices estimate for cosmological parameters inference, and on the impact of systematics due to baryonic effects. We present a machine learning approach to emulate the hydrodynamical simulation of intergalactic medium physics for the Lyman-$\alpha$ forest called LyAl-Net. The main goal of this work is to provide highly efficient and cheap simulations retaining interpretation abilities about the gas field level, and as a tool for other cosmological exploration. We use a neural network based on the U-net architecture, a variant of convolutional neural networks, to predict the neutral hydrogen physical properties, density, and temperature. We train the LyAl-Net model with the Horizon-noAGN simulation, though using only 9% of the volume. We also explore the resilience of the model through tests of a transfer learning framework using cosmological simulations containing different baryonic feedback. We test our results by analysing one and two-point statistics of emulated fields in different scenarios, as well as their stochastic properties. The ensemble average of the emulated Lyman-$\alpha$ forest absorption as a function of redshift lies within 2.5% of one derived from the full hydrodynamical simulation. The computation of individual fields from the dark matter density agrees well with regular physical regimes of cosmological fields. The results tested on IllustrisTNG100 showed a drastic improvement in the Lyman-$\alpha$ forest flux without arbitrary rescaling.

Read this paper on arXiv…

C. Boonkongkird, G. Lavaux, S. Peirani, et. al.
Mon, 3 Apr 23
48/53

Comments: N/A

APES: Approximate Posterior Ensemble Sampler [CEA]

Posted on by

http://arxiv.org/abs/2303.13667


This paper proposes a novel approach to generate samples from target distributions that are difficult to sample from using Markov Chain Monte Carlo (MCMC) methods. Traditional MCMC algorithms often face slow convergence due to the difficulty in finding proposals that suit the problem at hand. To address this issue, the paper introduces the Approximate Posterior Ensemble Sampler (APES) algorithm, which employs kernel density estimation and radial basis interpolation to create an adaptive proposal, leading to fast convergence of the chains. The APES algorithm’s scalability to higher dimensions makes it a practical solution for complex problems. The proposed method generates an approximate posterior probability that closely approximates the desired distribution and is easy to sample from, resulting in smaller autocorrelation times and a higher probability of acceptance by the chain. In this work, we compare the performance of the APES algorithm with the affine invariance ensemble sampler with the stretch move in various contexts, demonstrating the efficiency of the proposed method. For instance, on the Rosenbrock function, the APES presented an autocorrelation time 140 times smaller than the affine invariance ensemble sampler. The comparison showcases the effectiveness of the APES algorithm in generating samples from challenging distributions. This paper presents a practical solution to generating samples from complex distributions while addressing the challenge of finding suitable proposals. With new cosmological surveys set to deal with many new systematics, which will require many new nuisance parameters in the models, this method offers a practical solution for the upcoming era of cosmological analyses.

Read this paper on arXiv…

S. Vitenti and E. Barroso
Mon, 27 Mar 23
33/59

Comments: 15 pages, 6 figures, 7 tables

Optimal observational scheduling framework for binary and multiple stellar systems [SSA]

Posted on by

http://arxiv.org/abs/2301.04162


The optimal instant of observation of astrophysical phenomena for objects that vary on human time-sales is an important problem, as it bears on the cost-effective use of usually scarce observational facilities. In this paper we address this problem for the case of tight visual binary systems through a Bayesian framework based on the maximum entropy sampling principle. Our proposed information-driven methodology exploits the periodic structure of binary systems to provide a computationally efficient estimation of the probability distribution of the optimal observation time. We show the optimality of the proposed sampling methodology in the Bayes sense and its effectiveness through direct numerical experiments. We successfully apply our scheme to the study of two visual-spectroscopic binaries, and one purely astrometric triple hierarchical system. We note that our methodology can be applied to any time-evolving phenomena, a particularly interesting application in the era of dedicated surveys, where a definition of the cadence of observations can have a crucial impact on achieving the science goals.

Read this paper on arXiv…

M. Videla, R. Mendez, J. Silva, et. al.
Thu, 12 Jan 23
19/68

Comments: Accepted for publication to PASP. 23 pages, 2 Tables, 9 Figures, 2 Appendices

Low-Frequency Noise Mitigation and Bandgap Engineering using Seismic Metamaterials for Terrestrial Gravitational Wave Observatories [IMA]

Posted on by

http://arxiv.org/abs/2301.04325


Gravitational-wave now became one of the important observational methods for studying the Universe since its first detection. However, the ground-based observatories have an inherent barrier to their detection frequency band due to the seismic and gravity gradient noises nearby the perturbation of the surroundings. A recent intriguing development of artificial structures for media called metamaterial is opening a new branch of wave mechanics and its application in various fields, in particular, suggesting a novel way of mitigating noises by controlling the media structure for propagating waves. In this paper, we propose a novel framework for handling noises in ground-based gravitational wave detectors by using wave mechanics under metamaterial media. Specifically, we suggest an application of the bandgap engineering technique for mitigating the underground effects of acoustic noises resulting from the seismic vibration in the KAGRA gravitational wave observatory.

Read this paper on arXiv…

J. Oh
Thu, 12 Jan 23
42/68

Comments: 7pages, 5figures

Science Platforms for Heliophysics Data Analysis [IMA]

Posted on by

http://arxiv.org/abs/2301.00878


We recommend that NASA maintain and fund science platforms that enable interactive and scalable data analysis in order to maximize the scientific return of data collected from space-based instruments.

Read this paper on arXiv…

M. Bobra, W. Barnes, T. Chen, et. al.
Wed, 4 Jan 23
17/43

Comments: Heliophysics 2050 White Paper

Towards data-driven modeling and real-time prediction of solar flares and coronal mass ejections [IMA]

Posted on by

http://arxiv.org/abs/2212.14384


Modeling of transient events in the solar atmosphere requires the confluence of 3 critical elements: (1) model sophistication, (2) data availability, and (3) data assimilation. This white paper describes required advances that will enable statistical flare and CME forecasting (e.g. eruption probability and timing, estimation of strength, and CME details, such as speed and magnetic field orientation) similar to weather prediction on Earth.

Read this paper on arXiv…

M. Rempel, Y. Fan, M. Dikpati, et. al.
Mon, 2 Jan 23
14/44

Comments: Heliophysics 2050 White Paper

Detecting neutrinos in IceCube with Cherenkov light in the South Pole ice [IMA]

Posted on by

http://arxiv.org/abs/2212.12142


The IceCube Neutrino Observatory detects GeV-to-PeV+ neutrinos via the Cherenkov light produced by secondary charged particles from neutrino interactions with the South Pole ice. The detector consists of over 5000 spherical Digital Optical Modules (DOM), each deployed with a single downward-facing photomultiplier tube (PMT) and arrayed across 86 strings over a cubic-kilometer. IceCube has measured the astrophysical neutrino flux, searched for their origins, and constrained neutrino oscillation parameters and cross sections. These were made possible by an in-depth characterization of the glacial ice, which has been refined over time, and novel approaches in reconstructions that utilize fast approximations of Cherenkov yield expectations.
After over a decade of nearly continuous IceCube operation, the next generation of neutrino telescopes at the South Pole are taking shape. The IceCube Upgrade will add seven additional strings in a dense infill configuration. Multi-PMT OMs will be attached to each string, along with improved calibration devices and new sensor prototypes. Its denser OM and string spacing will extend sensitivity to lower neutrino energies and further constrain neutrino oscillation parameters. The calibration goals of the Upgrade will help guide the design and construction of IceCube Gen2, which will increase the effective volume by nearly an order of magnitude.

Read this paper on arXiv…

T. Yuan
Mon, 26 Dec 22
38/39

Comments: 5 pages, 5 figures, proceeding from the 11th International Workshop on Ring Imaging Cherenkov Detectors (RICH2022)

MFV approach to robust estimate of neutron lifetime [CL]

Posted on by

http://arxiv.org/abs/2212.05890


Aiming at evaluating the lifetime of the neutron, we introduce a novel statistical method to analyse the updated compilation of precise measurements including the 2022 dataset of Particle Data Group (PDG). Based on the minimization for the information loss principle, unlike the median statistics method, we apply the most frequent value (MFV) procedure to estimate the neutron lifetime, irrespective of the Gaussian or non-Gaussian distributions. Providing a more robust way, the calculated result of the MFV is $\tau_n=881.16^{+2.25}{-2.35}$ s with statistical bootstrap errors, while the result of median statistics is $\tau_n=881.5^{+5.5}{-3}$ s according to the binomial distribution. Using the different central estimates, we also construct the error distributions of neutron lifetime measurements and find the non-Gaussianity, which is still meaningful.

Read this paper on arXiv…

J. Zhang, S. Zhang, Z. Zhang, et. al.
Tue, 13 Dec 22
14/105

Comments: N/A

Measuring the fine structure constant on white dwarf surfaces; uncertainties from continuum placement variations [IMA]

Posted on by

http://arxiv.org/abs/2212.00434


Searches for variations of fundamental constants require accurate measurement errors. There are several potential sources of errors and quantifying each one accurately is essential. This paper addresses one source of uncertainty relating to measuring the fine structure constant on white dwarf surfaces. Detailed modelling of photospheric absorption lines requires knowing the underlying spectral continuum level. Here we describe the development of a fully automated, objective, and reproducible continuum estimation method, based on fitting cubic splines to carefully selected data regions. Example fits to the Hubble Space Telescope spectrum of the white dwarf G191-B2B are given. We carry out measurements of the fine structure constant using two continuum models. The results show that continuum placement variations result in small systematic shifts in the centroids of narrow photospheric absorption lines which impact significantly on fine structure constant measurements. This effect must therefore be included in the overall error budget of future measurements. Our results also suggest that continuum placement variations should be investigated in other contexts, including fine structure constant measurements in stars other than white dwarfs, quasar absorption line measurements of the fine structure constant, and quasar measurements of cosmological redshift drift.

Read this paper on arXiv…

C. Lee, J. Webb, D. Dougan, et. al.
Fri, 2 Dec 22
18/81

Comments: 10 pages, 4 figures. 4 additional files provided as supplementary material. Submitted to MNRAS 1 Dec 2022

Varying alpha, blinding, and bias in existing measurements [CEA]

Posted on by

http://arxiv.org/abs/2212.00791


The high resolution spectrograph ESPRESSO on the VLT allows measurements of fundamental constants at unprecedented precision and hence enables tests for spacetime variations predicted by some theories. In a series of recent papers, we developed optimal analysis procedures that both exposes and eliminates the subjectivity and bias in previous quasar absorption system measurements. In this paper we analyse the ESPRESSO spectrum of the absorption system at z_{abs}=1.15 towards the quasar HE0515-4414. Our goal here is not to provide a new unbiased measurement of fine structure constant, alpha, in this system (that will be done separately). Rather, it is to carefully examine the impact of blinding procedures applied in the recent analysis of the same data by Murphy (2022) and prior to that, in several other analyses. To do this we use supercomputer Monte Carlo AI calculations to generate a large number of independently constructed models of the absorption complex. Each model is obtained using AI-VPFIT, with alpha fixed until a “final” model is obtained, at which point alpha is then released as a free parameter for one final optimisation. The results show that the “measured” value of alpha is systematically biased towards the initially-fixed value i.e. this process produces meaningless measurements. The implication is straightforward: to avoid bias, all future measurements must include alpha as a free parameter from the beginning of the modelling process.

Read this paper on arXiv…

C. Lee, J. Webb, R. Carswell, et. al.
Fri, 2 Dec 22
47/81

Comments: 17 pages, 7 figures, and 5 tables. Submitted to MNRAS 1 Dec 2022

Cosmic-Ray Composition analysis at IceCube using Graph Neural Networks [HEAP]

Posted on by

http://arxiv.org/abs/2211.17198


The IceCube Neutrino Observatory is a multi-component detector embedded deep within the South-Pole Ice. This proceeding will discuss an analysis from an integrated operation of IceCube and its surface array, IceTop, to estimate cosmic-ray composition. The work will describe a novel graph neural network based approach for estimating the mass of primary cosmic rays, that takes advantage of signal-footprint information and reconstructed cosmic-ray air shower parameters. In addition, the work will also introduce new composition-sensitive parameters for improving the estimation of cosmic-ray composition, with the potential of improving our understanding of the high-energy muon content in cosmic-ray air showers.

Read this paper on arXiv…

P. Koundal
Thu, 1 Dec 22
26/85

Comments: N/A

Nested sampling statistical errors [IMA]

Posted on by

http://arxiv.org/abs/2211.03258


Nested sampling (NS) is a popular algorithm for Bayesian computation. We investigate statistical errors in NS both analytically and numerically. We show two analytic results. First, we show that the leading terms in Skilling’s expression using information theory match the leading terms in Keeton’s expression from an analysis of moments. This approximate agreement was previously only known numerically and was somewhat mysterious. Second, we show that the uncertainty in single NS runs approximately equals the standard deviation in repeated NS runs. Whilst intuitive, this was previously taken for granted. We close by investigating our results and their assumptions in several numerical examples, including cases in which NS uncertainties increase without bound.

Read this paper on arXiv…

A. Fowlie, Q. Li, H. Lv, et. al.
Tue, 8 Nov 22
16/79

Comments: 12 pages + appendices, 3 figures

A general method for goodness-of-fit tests for arbitrary multivariate models [CL]

Posted on by

http://arxiv.org/abs/2211.03478


Goodness-of-fit tests are often used in data analysis to test the agreement of a model to a set of data. Out of the box tests that can target any proposed distribution model are only available in the univariate case. In this note I discuss how to build a goodness-of-fit test for arbitrary multivariate distributions or multivariate data generation models. The resulting tests perform an unbinned analysis and do not need any trials factor or look-elsewhere correction since the multivariate data can be analyzed all at once. The proposed distribution or generative model is used to transform the data to an uncorrelated space where the test is developed. Depending on the complexity of the model, it is possible to perform the transformation analytically or numerically with the help of a Normalizing Flow algorithm.

Read this paper on arXiv…

L. Shtembari
Tue, 8 Nov 22
79/79

Comments: N/A

A robust estimator of mutual information for deep learning interpretability [CL]

Posted on by

http://arxiv.org/abs/2211.00024


We develop the use of mutual information (MI), a well-established metric in information theory, to interpret the inner workings of deep learning models. To accurately estimate MI from a finite number of samples, we present GMM-MI (pronounced $“$Jimmie$”$), an algorithm based on Gaussian mixture models that can be applied to both discrete and continuous settings. GMM-MI is computationally efficient, robust to the choice of hyperparameters and provides the uncertainty on the MI estimate due to the finite sample size. We extensively validate GMM-MI on toy data for which the ground truth MI is known, comparing its performance against established mutual information estimators. We then demonstrate the use of our MI estimator in the context of representation learning, working with synthetic data and physical datasets describing highly non-linear processes. We train deep learning models to encode high-dimensional data within a meaningful compressed (latent) representation, and use GMM-MI to quantify both the level of disentanglement between the latent variables, and their association with relevant physical quantities, thus unlocking the interpretability of the latent representation. We make GMM-MI publicly available.

Read this paper on arXiv…

D. Piras, H. Peiris, A. Pontzen, et. al.
Wed, 2 Nov 22
21/67

Comments: 13 pages, 7 figures, comments welcome. GMM-MI available at this https URL

Learning to Detect Interesting Anomalies [CL]

Posted on by

http://arxiv.org/abs/2210.16334


Anomaly detection algorithms are typically applied to static, unchanging, data features hand-crafted by the user. But how does a user systematically craft good features for anomalies that have never been seen? Here we couple deep learning with active learning — in which an Oracle iteratively labels small amounts of data selected algorithmically over a series of rounds — to automatically and dynamically improve the data features for efficient outlier detection. This approach, AHUNT, shows excellent performance on MNIST, CIFAR10, and Galaxy-DESI data, significantly outperforming both standard anomaly detection and active learning algorithms with static feature spaces. Beyond improved performance, AHUNT also allows the number of anomaly classes to grow organically in response to Oracle’s evaluations. Extensive ablation studies explore the impact of Oracle question selection strategy and loss function on performance. We illustrate how the dynamic anomaly class taxonomy represents another step towards fully personalized rankings of different anomaly classes that reflect a user’s interests, allowing the algorithm to learn to ignore statistically significant but uninteresting outliers (e.g., noise). This should prove useful in the era of massive astronomical datasets serving diverse sets of users who can only review a tiny subset of the incoming data.

Read this paper on arXiv…

A. Sadr, B. Bassett and E. Sekyi
Tue, 1 Nov 22
39/100

Comments: 10 pages, 7 figures

Sequential hypothesis testing for Axion Haloscopes [CL]

Posted on by

http://arxiv.org/abs/2210.16095


The goal of this paper is to introduce a novel likelihood-based inferential framework for axion haloscopes which is valid under the commonly applied “rescanning” protocol. The proposed method enjoys short data acquisition times and a simple tuning of the detector configuration. Local statistical significance and power are computed analytically, avoiding the need of burdensome simulations. Adequate corrections for the look-elsewhere effect are also discussed. The performance of our inferential strategy is compared with that of a simple method which exploits the geometric probability of rescan. Finally, we exemplify the method with an application to a HAYSTAC type axion haloscope.

Read this paper on arXiv…

A. Rosso, S. Algeri and J. Conrad
Mon, 31 Oct 22
48/60

Comments: 14 pages, 12 figures

Stokes inversion techniques with neural networks: analysis of uncertainty in parameter estimation [SSA]

Posted on by

http://arxiv.org/abs/2210.14933


Magnetic fields are responsible for a multitude of Solar phenomena, including such destructive events as solar flares and coronal mass ejections, with the number of such events rising as we approach the peak of the 11-year solar cycle, in approximately 2025. High-precision spectropolarimetric observations are necessary to understand the variability of the Sun. The field of quantitative inference of magnetic field vectors and related solar atmospheric parameters from such observations has long been investigated. In recent years, very sophisticated codes for spectropolarimetric observations have been developed. Over the past two decades, neural networks have been shown to be a fast and accurate alternative to classic inversion technique methods. However, most of these codes can be used to obtain point estimates of the parameters, so ambiguities, the degeneracies, and the uncertainties of each parameter remain uncovered. In this paper, we provide end-to-end inversion codes based on the simple Milne-Eddington model of the stellar atmosphere and deep neural networks to both parameter estimation and their uncertainty intervals. The proposed framework is designed in such a way that it can be expanded and adapted to other atmospheric models or combinations of them. Additional information can also be incorporated directly into the model. It is demonstrated that the proposed architecture provides high accuracy of results, including a reliable uncertainty estimation, even in the multidimensional case. The models are tested using simulation and real data samples.

Read this paper on arXiv…

L. Mistryukova, A. Plotnikov, A. Khizhik, et. al.
Fri, 28 Oct 22
35/56

Comments: 17 pages with 7 figures and 3 tables, submitted to Solar Physics

GraphNeT: Graph neural networks for neutrino telescope event reconstruction [IMA]

Posted on by

http://arxiv.org/abs/2210.12194


GraphNeT is an open-source python framework aimed at providing high quality, user friendly, end-to-end functionality to perform reconstruction tasks at neutrino telescopes using graph neural networks (GNNs). GraphNeT makes it fast and easy to train complex models that can provide event reconstruction with state-of-the-art performance, for arbitrary detector configurations, with inference times that are orders of magnitude faster than traditional reconstruction techniques. GNNs from GraphNeT are flexible enough to be applied to data from all neutrino telescopes, including future projects such as IceCube extensions or P-ONE. This means that GNN-based reconstruction can be used to provide state-of-the-art performance on most reconstruction tasks in neutrino telescopes, at real-time event rates, across experiments and physics analyses, with vast potential impact for neutrino and astro-particle physics.

Read this paper on arXiv…

A. Søgaard, R. Ørsøe, L. Bozianu, et. al.
Tue, 25 Oct 22
98/111

Comments: 6 pages, 1 figure. Code can be found at this https URL . Submitted to the Journal of Open Source Software (JOSS)

Characterization of the Thermospheric Mean Winds and Circulation during Solstice using ICON/MIGHTI Observations [CL]

Posted on by

http://arxiv.org/abs/2210.09407


Using the horizontal neutral wind observations from the MIGHTI instrument onboard NASA’s ICON (Ionospheric Connection Explorer) spacecraft with continuous coverage, we determine the climatology of the mean zonal and meridional winds and the associated mean circulation at low- to middle latitudes ($10^\circ$S-40$^{\circ}$N) for Northern Hemisphere {summer} solstice conditions between 90 km and 200 km altitudes, specifically on 20 June 2020 solstice as well as for a one-month period from 8 June-7 July 2020 {and for Northern winter season from 16 December 2019-31 January 2020, which spans a 47-day period, providing full local time coverage}. The data are averaged within appropriate altitude, longitude, latitude, solar zenith angle, and local time bins to produce mean wind distributions. The geographical distributions and local time variations of the mean horizontal circulation are evaluated. The instantaneous horizontal winds exhibit a significant degree of spatiotemporal variability often exceeding $\pm 150 $ m s$^{-1}$. The daily averaged zonal mean winds demonstrate day-to-day variability. Eastward zonal winds and northward (winter-to-summer) meridional winds are prevalent in the lower thermosphere, which provides indirect observational evidence of the eastward momentum deposition by small-scale gravity waves. The mean neutral winds and circulation exhibit smaller scale structures in the lower thermosphere (90-120 km), while they are more homogeneous in the upper thermosphere, indicating the increasingly dissipative nature of the thermosphere. The mean wind and circulation patterns inferred from ICON/MIGHTI measurements can be used to constrain and validate general circulation models, as well as input for numerical wave models.

Read this paper on arXiv…

E. Yiğit, M. Dhadly, A. Medvedev, et. al.
Wed, 19 Oct 22
24/87

Comments: Accepted for publication in Journal of Geophysical Research – Space Physics

Estimation of the number of counts on a particle counter detector with full time resolution [CL]

Posted on by

http://arxiv.org/abs/2210.09005


We present a general method for estimating the number of particles impinging on a segmented counter or, in general, on a counter with sub-units. We account for unresolved particles, i.e., the effect of two or more particles hitting the same sub-unit almost simultaneously. To achieve full time resolution we account for the dead time that occurs after the first time-bin of a particle signal. This general counting method can be applied to counting muons in existing detectors like the Underground Muon Detector of the Pierre Auger Observatory. We therefore use the latter as a study case to test the performance of our method and to compare it to other methods from literature. Our method proves to perform with little bias, and also provides an estimate of the number of particles as a function of time (as seen by the detector) to a single time-bin resolution. In this context, the new method can be useful for reconstructing parameters sensitive to cosmic ray mass, which are key to unveiling the origin of cosmic rays.

Read this paper on arXiv…

F. Gesualdi and A. Supanitsky
Tue, 18 Oct 22
45/99

Comments: Accepted for publication in EPJ C

Partition function approach to non-Gaussian likelihoods: Formalism and expansions for weakly non-Gaussian cosmological inference [CEA]

Posted on by

http://arxiv.org/abs/2210.03138


Non-Gaussian likelihoods, ubiquitous throughout cosmology, are a direct consequence of nonlinearities in the physical model. Their treatment requires Monte-Carlo Markov-chain or more advanced sampling methods for the determination of confidence contours. As an alternative, we construct canonical partition functions as Laplace-transforms of the Bayesian evidence, from which MCMC-methods would sample microstates. Cumulants of order $n$ of the posterior distribution follow by direct $n$-fold differentiation of the logarithmic partition function, recovering the classic Fisher-matrix formalism at second order. We connect this approach for weakly non-Gaussianities to the DALI- and Gram-Charlier expansions and demonstrate the validity with a supernova-likelihood on the cosmological parameters $\Omega_m$ and $w$. We comment on extensions of the canonical partition function to include kinetic energies in order to bridge to Hamilton Monte-Carlo sampling, and on ensemble Markov-chain methods, as they would result from transitioning to macrocanonical partition functions depending on a chemical potential. Lastly we demonstrate the relationship of the partition function approach to the Cram\’er-Rao boundary and to information entropies.

Read this paper on arXiv…

L. Röver, L. Bartels and B. Schäfer
Mon, 10 Oct 22
32/59

Comments: 12 pages 2 figures

A comparative study on different background estimation methods for extensive air shower arrays [IMA]

Posted on by

http://arxiv.org/abs/2210.00004


By applying four different methods including equi-zenith angle method, surrounding window method, direct integration method, and time-swapping method, the number of the background events is calculated. Based on simulation samples, the statistical significance of the excess signal from different background methods is determined. After that, we discuss the limits and the applicability of the four background method under different conditions. Under the detector stability assumption with signal, the results from above four methods are consistent within 1{\sigma} level. On no signal condition, when the acceptance of the detector changes with both space and time, the surrounding window method is most stable and hardly affected. In this acceptance assumption, we find that the background estimation in the direct integration method is sensitive to the selection of time integration window, and the time window of 4 hour is more applicable, which can reduce the impact on the background estimation to some extent.

Read this paper on arXiv…

Y. Wang, M. Zha, S. Hu, et. al.
Tue, 4 Oct 22
14/71

Comments: 19 pages

Six textbook mistakes in data analysis [CL]

Posted on by

http://arxiv.org/abs/2209.09073


This article discusses a number of incorrect statements appearing in textbooks on data analysis, machine learning, or computational methods; the common theme in all these cases is the relevance and application of statistics to the study of scientific or engineering data; these mistakes are also quite prevalent in the research literature. Crucially, we do not address errors made by an individual author, focusing instead on mistakes that are widespread in the introductory literature. After some background on frequentist and Bayesian linear regression, we turn to our six paradigmatic cases, providing in each instance a specific example of the textbook mistake, pointers to the specialist literature where the topic is handled properly, along with a correction that summarizes the salient points. The mistakes (and corrections) are broadly relevant to any technical setting where statistical techniques are used to draw practical conclusions, ranging from topics introduced in an elementary course on experimental measurements all the way to more involved approaches to regression.

Read this paper on arXiv…

A. Gezerlis and M. Williams
Tue, 20 Sep 22
13/81

Comments: 15 pages, 7 figures

Lessons Learned from the Two Largest Galaxy Morphological Classification Catalogues built by Convolutional Neural Networks [GA]

Posted on by

http://arxiv.org/abs/2209.06897


We compare the two largest galaxy morphology catalogues, which separate early and late type galaxies at intermediate redshift. The two catalogues were built by applying supervised deep learning (convolutional neural networks, CNNs) to the Dark Energy Survey data down to a magnitude limit of $\sim$21 mag. The methodologies used for the construction of the catalogues include differences such as the cutout sizes, the labels used for training, and the input to the CNN – monochromatic images versus $gri$-band normalized images. In addition, one catalogue is trained using bright galaxies observed with DES ($i<18$), while the other is trained with bright galaxies ($r<17.5$) and `emulated’ galaxies up to $r$-band magnitude $22.5$. Despite the different approaches, the agreement between the two catalogues is excellent up to $i<19$, demonstrating that CNN predictions are reliable for samples at least one magnitude fainter than the training sample limit. It also shows that morphological classifications based on monochromatic images are comparable to those based on $gri$-band images, at least in the bright regime. At fainter magnitudes, $i>19$, the overall agreement is good ($\sim$95\%), but is mostly driven by the large spiral fraction in the two catalogues. In contrast, the agreement within the elliptical population is not as good, especially at faint magnitudes. By studying the mismatched cases we are able to identify lenticular galaxies (at least up to $i<19$), which are difficult to distinguish using standard classification approaches. The synergy of both catalogues provides an unique opportunity to select a population of unusual galaxies.

Read this paper on arXiv…

T. Cheng, H. Sánchez, J. Vega-Ferrero, et. al.
Fri, 16 Sep 22
46/84

Comments: 17 pages, 14 figures (1 appendix for galaxy examples including 3 figures)

Evaluating the efficacy of sonification for signal detection in univariate, evenly sampled light curves using astronify [IMA]

Posted on by

http://arxiv.org/abs/2209.04465


Sonification is the technique of representing data with sound, with potential applications in astronomy research for aiding discovery and accessibility. Several astronomy-focused sonification tools have been developed; however, efficacy testing is extremely limited. We performed testing of astronify, a prototype tool for sonification functionality within the Barbara A. Mikulski Archive for Space Telescopes (MAST). We created synthetic light curves containing zero, one, or two transit-like signals with a range of signal-to-noise ratios (SNRs=3-100) and applied the default mapping of brightness to pitch. We performed remote testing, asking participants to count signals when presented with light curves as a sonification, visual plot, or combination of both. We obtained 192 responses, of which 118 self-classified as experts in astronomy and data analysis. For high SNRs (=30 and 100), experts and non-experts performed well with sonified data (85-100% successful signal counting). At low SNRs (=3 and 5) both groups were consistent with guessing with sonifications. At medium SNRs (=7 and 10), experts performed no better than non-experts with sonifications but significantly better (factor of ~2-3) with visuals. We infer that sonification training, like that experienced by experts for visual data inspection, will be important if this sonification method is to be useful for moderate SNR signal detection within astronomical archives and broader research. Nonetheless, we show that even a very simple, and non-optimised, sonification approach allows users to identify high SNR signals. A more optimised approach, for which we present ideas, would likely yield higher success for lower SNR signals.

Read this paper on arXiv…

J. Brown, C. Harrison, A. Zanella, et. al.
Tue, 13 Sep 22
48/85

Comments: Accepted for publication in MNRAS (10 pages, 5 figures). Sonifications of Figure 1 (4 audio files) and Figure 5 (2 movie files) are available in the ancillary files folder. These, plus all other data products associated with this article are also available at: this https URL

A characterization method for low-frequency environmental noise in LIGO [IMA]

Posted on by

http://arxiv.org/abs/2209.04452


We present a method to characterize the noise in ground-based gravitational-wave observatories such as the Laser Gravitational-Wave Observatory (LIGO). This method uses linear regression algorithms such as the least absolute shrinkage and selection operator (LASSO) to identify noise sources and analyzes the detector output versus noise witness sensors to quantify the coupling of such noise. Our method can be implemented with currently available resources at LIGO, which avoids extra coding or direct experimentation at the LIGO sites. We present two examples to validate and estimate the coupling of elevated ground motion at frequencies below 10 Hz with noise in the detector output.

Read this paper on arXiv…

G. Valdes, A. Hines, A. Nelson, et. al.
Tue, 13 Sep 22
61/85

Comments: N/A

Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube [CL]

Posted on by

http://arxiv.org/abs/2209.03042


IceCube, a cubic-kilometer array of optical sensors built to detect atmospheric and astrophysical neutrinos between 1 GeV and 1 PeV, is deployed 1.45 km to 2.45 km below the surface of the ice sheet at the South Pole. The classification and reconstruction of events from the in-ice detectors play a central role in the analysis of data from IceCube. Reconstructing and classifying events is a challenge due to the irregular detector geometry, inhomogeneous scattering and absorption of light in the ice and, below 100 GeV, the relatively low number of signal photons produced per event. To address this challenge, it is possible to represent IceCube events as point cloud graphs and use a Graph Neural Network (GNN) as the classification and reconstruction method. The GNN is capable of distinguishing neutrino events from cosmic-ray backgrounds, classifying different neutrino event types, and reconstructing the deposited energy, direction and interaction vertex. Based on simulation, we provide a comparison in the 1-100 GeV energy range to the current state-of-the-art maximum likelihood techniques used in current IceCube analyses, including the effects of known systematic uncertainties. For neutrino event classification, the GNN increases the signal efficiency by 18% at a fixed false positive rate (FPR), compared to current IceCube methods. Alternatively, the GNN offers a reduction of the FPR by over a factor 8 (to below half a percent) at a fixed signal efficiency. For the reconstruction of energy, direction, and interaction vertex, the resolution improves by an average of 13%-20% compared to current maximum likelihood techniques in the energy range of 1-30 GeV. The GNN, when run on a GPU, is capable of processing IceCube events at a rate nearly double of the median IceCube trigger rate of 2.7 kHz, which opens the possibility of using low energy neutrinos in online searches for transient events.

Read this paper on arXiv…

R. Abbasi, M. Ackermann, J. Adams, et. al.
Thu, 8 Sep 22
49/77

Comments: Prepared for submission to JINST

Virgo: Scalable Unsupervised Classification of Cosmological Shock Waves [IMA]

Posted on by

http://arxiv.org/abs/2208.06859


Cosmological shock waves are essential to understanding the formation of cosmological structures. To study them, scientists run computationally expensive high-resolution 3D hydrodynamic simulations. Interpreting the simulation results is challenging because the resulting data sets are enormous, and the shock wave surfaces are hard to separate and classify due to their complex morphologies and multiple shock fronts intersecting. We introduce a novel pipeline, Virgo, combining physical motivation, scalability, and probabilistic robustness to tackle this unsolved unsupervised classification problem. To this end, we employ kernel principal component analysis with low-rank matrix approximations to denoise data sets of shocked particles and create labeled subsets. We perform supervised classification to recover full data resolution with stochastic variational deep kernel learning. We evaluate on three state-of-the-art data sets with varying complexity and achieve good results. The proposed pipeline runs automatically, has only a few hyperparameters, and performs well on all tested data sets. Our results are promising for large-scale applications, and we highlight now enabled future scientific work.

Read this paper on arXiv…

M. Lamparth, L. Böss, U. Steinwandel, et. al.
Tue, 16 Aug 22
46/74

Comments: N/A

The entropy of galaxy spectra: How much information is encoded? [GA]

Posted on by

http://arxiv.org/abs/2208.05489


This paper approaches the inverse problem of extracting the stellar population content of galaxy spectra from a basic standpoint based on information theory. By interpreting spectra as probability distribution functions, we find that galaxy spectra have high entropy, caused by the high correlatedness in wavelength space. The highest variation in entropy is unsurprisingly found in regions that have been well studied for decades with the conventional approach. Therefore, we target a set of six spectral regions that show the highest variation in entropy – the 4,000 Angstrom break being the most informative one. As a test case with real data, we measure the entropy of a set of high quality spectra from the Sloan Digital Sky Survey, and contrast entropy-based results with the traditional method based on line strengths. The data are classified into star-forming (SF), quiescent (Q) and AGN galaxies, and show – independently of any physical model – that AGN spectra represent a transition between SF and Q galaxies, with SF galaxies featuring a more diverse variation in entropy. The high level of entanglement complicates the determination of population parameters in a robust, unbiased way, and affect traditional methods that compare models with observations, as well as machine learning and deep learning algorithms that rely on the statistical properties of the data to assess the variations among spectra. Therefore, caution must be exercised when retrieving detailed population parameters or even star formation histories from galaxy spectra.

Read this paper on arXiv…

I. Ferreras, O. Lahav, R. Somerville, et. al.
Fri, 12 Aug 22
44/48

Comments: 11 pages, 12 figures. Comments welcome

Likelihood-free Inference with Mixture Density Network [CEA]

Posted on by

http://arxiv.org/abs/2207.00185


In this work, we propose the mixture density network (MDN) to estimate cosmological parameters. We test the MDN method by constraining parameters of the $\Lambda$CDM and $w$CDM models using Type-Ia supernovae and power spectra of the cosmic microwave background. We find that the MDN method can achieve the same level of accuracy as the Markov Chain Monte Carlo method, with a slight difference of $\mathcal{O}(10^{-2}\sigma)$. Furthermore, the MDN method can provide accurate parameter estimates with $\mathcal{O}(10^3)$ forward simulation samples, which is useful for complex and resource-consuming cosmological models. This method can process either one data set or multiple data sets to achieve joint constraints on parameters, extendable for any parameter estimation of complicated models in a wider scientific field. Thus, MDN provides an alternative way for likelihood-free inference of parameters.

Read this paper on arXiv…

G. Wang, C. Cheng, Y. Ma, et. al.
Mon, 4 Jul 22
49/62

Comments: 17 pages, 4 tables, 15 figures, accepted by the Astrophysical Journal Supplement Series

Revealing the Milky Way's Most Recent Major Merger with a Gaia EDR3 Catalog of Machine-Learned Line-of-Sight Velocities [GA]

Posted on by

http://arxiv.org/abs/2205.12278


Machine learning can play a powerful role in inferring missing line-of-sight velocities from astrometry in surveys such as Gaia. In this paper, we apply a neural network to Gaia Early Data Release 3 (EDR3) and obtain line-of-sight velocities and associated uncertainties for ~92 million stars. The network, which takes as input a star’s parallax, angular coordinates, and proper motions, is trained and validated on ~6.4 million stars in Gaia with complete phase-space information. The network’s uncertainty on its velocity prediction is a key aspect of its design; by properly convolving these uncertainties with the inferred velocities, we obtain accurate stellar kinematic distributions. As a first science application, we use the new network-completed catalog to identify candidate stars that belong to the Milky Way’s most recent major merger, Gaia-Sausage-Enceladus (GSE). We present the kinematic, energy, angular momentum, and spatial distributions of the ~450,000 GSE candidates in this sample, and also study the chemical abundances of those with cross matches to GALAH and APOGEE. The network’s predictive power will only continue to improve with future Gaia data releases as the training set of stars with complete phase-space information grows. This work provides a first demonstration of how to use machine learning to exploit high-dimensional correlations on data to infer line-of-sight velocities, and offers a template for how to train, validate and apply such a neural network when complete observational data is not available.

Read this paper on arXiv…

A. Dropulic, H. Liu, B. Ostdiek, et. al.
Thu, 26 May 22
36/56

Comments: 18 pages, 11 figures

A method for approximating optimal statistical significances with machine-learned likelihoods [CL]

Posted on by

http://arxiv.org/abs/2205.05952


Machine-learning techniques have become fundamental in high-energy physics and, for new physics searches, it is crucial to know their performance in terms of experimental sensitivity, understood as the statistical significance of the signal-plus-background hypothesis over the background-only one. We present here a new method that combines the power of current machine-learning techniques to face high-dimensional data with the likelihood-based inference tests used in traditional analyses, which allows us to estimate the sensitivity for both discovery and exclusion limits through a single parameter of interest, the signal strength. Based on supervised learning techniques, it can perform well also with high-dimensional data, when traditional techniques cannot. We apply the method to a toy model first, so we can explore its potential, and then to a LHC study of new physics particles in dijet final states. Considering as the optimal statistical significance the one we would obtain if the true generative functions were known, we show that our method provides a better approximation than the usual naive counting experimental results.

Read this paper on arXiv…

E. Arganda, X. Marcano, V. Lozano, et. al.
Fri, 13 May 22
59/64

Comments: 22 pages, 8 figures. Comments welcome!

An intuition for physicists: information gain from experiments [CL]

Posted on by

http://arxiv.org/abs/2205.00009


How much one has learned from an experiment is quantifiable by the information gain, also known as the Kullback-Leibler divergence. The narrowing of the posterior parameter distribution $P(\theta|D)$ compared with the prior parameter distribution $\pi(\theta)$, is quantified in units of bits, as: $ D_{\mathrm{KL}}(P|\pi)=\int\log_{2}\left(\frac{P(\theta|D)}{\pi(\theta)}\right)\,P(\theta|D)\,d\theta $. This research note gives an intuition what one bit of information gain means. It corresponds to a Gaussian shrinking its standard deviation by a factor of three.

Read this paper on arXiv…

J. Buchner
Tue, 3 May 22
34/82

Comments: Accepted to RNAAS

Gravitational Waves from Double White Dwarfs as probes of the Milky Way [GA]

Posted on by

http://arxiv.org/abs/2204.07349


Future gravitational wave detectors, such as the Laser Interferometer Space Antenna (LISA), will be able to resolve a significant number of the ultra compact stellar-mass binaries in our own Galaxy and its neighborhood. These will be mostly double white dwarf (DWD) binaries, and their underlying population characteristics can be directly correlated to the different properties of the Galaxy. In particular, with LISA we will be able to resolve $\sim\mathcal{O}(10^4)$ binaries, while the rest will generate a confusion foreground signal. Both categories can be used to address a number of astrophysical questions. Analogously to how the total electromagnetic radiation emitted by a galaxy can be related to the underlying total stellar mass, in this work we propose a framework to infer the same quantity by investigating the spectral shape and amplitude of the confusion foreground signal. For a fixed DWD evolution model, we retrieve percentage-level relative errors on the total stellar mass, which improves for increasing values of the mass. At the same time, we find that variations in the Miky Way shape, at a fixed mass and at scale heights smaller than 500~pc, are not distinguishable based on the shape of stochastic signal alone. Finally, we utilize the catalogue of resolvable sources to probe the characteristics of the underlying population of DWD binaries. We show that the DWD frequency, coalescence time and chirp mass (up to $<0.7\,$M$_\odot$) distributions can be reconstructed from LISA data with no bias.

Read this paper on arXiv…

M. Georgousi, N. Karnesis, V. Korol, et. al.
Mon, 18 Apr 22
20/34

Comments: 13 pages, 9 figures

Quantification of high dimensional non-Gaussianities and its implication to Fisher analysis in cosmology [CEA]

Posted on by

http://arxiv.org/abs/2204.05435


It is well known that the power spectrum is not able to fully characterize the statistical properties of non-Gaussian density fields. Recently, many different statistics have been proposed to extract information from non-Gaussian cosmological fields that perform better than the power spectrum. The Fisher matrix formalism is commonly used to quantify the accuracy with which a given statistic can constrain the value of the cosmological parameters. However, these calculations typically rely on the assumption that the likelihood of the considered statistic follows a multivariate Gaussian distribution. In this work we follow Sellentin & Heavens (2017) and use two different statistical tests to identify non-Gaussianities in different statistics such as the power spectrum, bispectrum, marked power spectrum, and wavelet scatering transform (WST). We remove the non-Gaussian components of the different statistics and perform Fisher matrix calculations with the \textit{Gaussianized} statistics using Quijote simulations. We show that constraints on the parameters can change by a factor of $\sim 2$ in some cases. We show with simple examples how statistics that do not follow a multivariate Gaussian distribution can achieve artificially tight bounds on the cosmological parameters when using the Fisher matrix formalism. We think that the non-Gaussian tests used in this work represent a powerful tool to quantify the robustness of Fisher matrix calculations and their underlying assumptions. We release the code used to compute the power spectra, bispectra, and WST that can be run on both CPUs and GPUs.

Read this paper on arXiv…

C. Park, E. Allys, F. Villaescusa-Navarro, et. al.
Wed, 13 Apr 22
19/73

Comments: 24 pages, 6 figures

Bayesian parameter-estimation of Galactic binaries in LISA data with Gaussian Process Regression [IMA]

Posted on by

http://arxiv.org/abs/2204.04467


The Laser Interferometer Space Antenna (LISA), which is currently under construction, is designed to measure gravitational wave signals in the milli-Hertz frequency band. It is expected that tens of millions of Galactic binaries will be the dominant sources of observed gravitational waves. The Galactic binaries producing signals at mHz frequency range emit quasi monochromatic gravitational waves, which will be constantly measured by LISA. To resolve as many Galactic binaries as possible is a central challenge of the upcoming LISA data set analysis. Although it is estimated that tens of thousands of these overlapping gravitational wave signals are resolvable, and the rest blurs into a galactic foreground noise; extracting tens of thousands of signals using Bayesian approaches is still computationally expensive. We developed a new end-to-end pipeline using Gaussian Process Regression to model the log-likelihood function in order to rapidly compute Bayesian posterior distributions. Using the pipeline we are able to solve the Lisa Data Challange (LDC) 1-3 consisting of noisy data as well as additional challenges with overlapping signals and particularly faint signals.

Read this paper on arXiv…

S. Strub, L. Ferraioli, C. Schmelzbach, et. al.
Tue, 12 Apr 22
25/87

Comments: 12 pages, 10 figures

Using Kernel-Based Statistical Distance to Study the Dynamics of Charged Particle Beams in Particle-Based Simulation Codes [CL]

Posted on by

http://arxiv.org/abs/2204.04275


Measures of discrepancy between probability distributions (statistical distance) are widely used in the fields of artificial intelligence and machine learning. We describe how certain measures of statistical distance can be implemented as numerical diagnostics for simulations involving charged-particle beams. Related measures of statistical dependence are also described. The resulting diagnostics provide sensitive measures of dynamical processes important for beams in nonlinear or high-intensity systems, which are otherwise difficult to characterize. The focus is on kernel-based methods such as Maximum Mean Discrepancy, which have a well-developed mathematical foundation and reasonable computational complexity. Several benchmark problems and examples involving intense beams are discussed. While the focus is on charged-particle beams, these methods may also be applied to other many-body systems such as plasmas or gravitational systems.

Read this paper on arXiv…

C. Mitchell, R. Ryne and K. Hwang
Tue, 12 Apr 22
87/87

Comments: N/A

Escaping the maze: a statistical sub-grid model for cloud-scale density structures in the interstellar medium [GA]

Posted on by

http://arxiv.org/abs/2204.02053


The interstellar medium (ISM) is a turbulent, highly structured multi-phase medium. State-of-the-art cosmological simulations of the formation of galactic discs usually lack the resolution to accurately resolve those multi-phase structures. However, small-scale density structures play an important role in the life cycle of the ISM, and determine the fraction of cold, dense gas, the amount of star formation and the amount of radiation and momentum leakage from cloud-embedded sources. Here, we derive a $statistical\, model$ to calculate the unresolved small-scale ISM density structure from coarse-grained, volume-averaged quantities such as the $gas\, clumping\, factor$, $\mathcal{C}$, and mean density $\left<\rho\right>V$. Assuming that the large-scale ISM density is statistically isotropic, we derive a relation between the three-dimensional clumping factor, $\mathcal{C}\rho$, and the clumping factor of the $4\pi$ column density distribution on the cloud surface, $\mathcal{C}\Sigma$, and find $\mathcal{C}\Sigma=\mathcal{C}_\rho^{2/3}$. Applying our model to calculate the covering fraction, i.e., the $4\pi$ sky distribution of optically thick sight-lines around sources inside interstellar gas clouds, we demonstrate that small-scale density structures lead to significant differences at fixed physical ISM density. Our model predicts that gas clumping increases the covering fraction by up to 30 per cent at low ISM densities compared to a uniform medium. On the other hand, at larger ISM densities, gas clumping suppresses the covering fraction and leads to increased scatter such that covering fractions can span a range from 20 to 100 per cent at fixed ISM density. All data and example code is publicly available at GitHub.

Read this paper on arXiv…

T. Buck, C. Pfrommer, P. Girichidis, et. al.
Wed, 6 Apr 22
67/68

Comments: 16 pages with 8 figures, 14 pages main text with 7 figures, 1 page references, 1 page appendix with 1 figure, accepted by MNRAS on April 1st 2022

Identifying and diagnosing coherent associations and causalities between multi-channels of the gravitational wave detector [IMA]

Posted on by

http://arxiv.org/abs/2204.00370


The gravitational-wave detector is a very complicated and sensitive collection of advanced instruments, which is influenced not only by the mutual interaction between mechanical/electronics systems but also by the surrounding environment. Thus, it is necessary to categorize and reduce noises from many channels interconnected by such instruments and environment for achieving the detection of gravitational waves because it enhances to increase of a signal-to-noise ratio and reduces false alarms from coincident loud events. For this reason, it is of great importance to identify some coherent associations between complicated channels. This study presents a way of identifying (non-) linear couplings between interconnected channels by using some correlation coefficients, which are applied to practical issues such as noises by hardware injection test, lightning strokes, and air compressor vibrations gravitational-wave detector.

Read this paper on arXiv…

P. Jung, S. Oh, Y. Kim, et. al.
Mon, 4 Apr 22
37/50

Comments: 10 pages, 8 figures, and 2 tables

Transverse Vector Decomposition Method for Analytical Inversion of Exoplanet Transit Spectra [EPA]

Posted on by

http://arxiv.org/abs/2203.06299


We develop a new method for analytical inversion of binned exoplanet transit spectra and for retrieval of planet parameters. The method has a geometrical interpretation and treats each observed spectrum as a single vector $\vec r$ in the multidimensional spectral space of observed bin values. We decompose the observed $\vec{r}$ into a wavelength-independent component $\vec{r}\parallel$ corresponding to the spectral mean across all observed bins, and a transverse component $\vec{r}\perp$ which is wavelength-dependent and contains the relevant information about the atmospheric chemistry. The method allows us to extract, without any prior assumptions or additional information, the relative mass (or volume) mixing ratios of the absorbers in the atmosphere, the scale height to stellar radius ratio, $H/R_S$, and the atmospheric temperature. The method is illustrated and validated with several examples of increasing complexity.

Read this paper on arXiv…

K. Matchev, K. Matcheva and A. Roman
Tue, 15 Mar 22
76/108

Comments: Submitted to AAS Journals, 22 pages, 9 figures

Low Energy Event Reconstruction in IceCube DeepCore [CL]

Posted on by

http://arxiv.org/abs/2203.02303


The reconstruction of event-level information, such as the direction or energy of a neutrino interacting in IceCube DeepCore, is a crucial ingredient to many physics analyses. Algorithms to extract this high level information from the detector’s raw data have been successfully developed and used for high energy events. In this work, we address unique challenges associated with the reconstruction of lower energy events in the range of a few to hundreds of GeV and present two separate, state-of-the-art algorithms. One algorithm focuses on the fast directional reconstruction of events based on unscattered light. The second algorithm is a likelihood-based multipurpose reconstruction offering superior resolutions, at the expense of larger computational cost.

Read this paper on arXiv…

R. Abbasi, M. Ackermann, J. Adams, et. al.
Mon, 7 Mar 22
19/64

Comments: N/A

SiGMa-Net: Deep learning network to distinguish binary black hole signals from short-duration noise transients [CL]

Posted on by

http://arxiv.org/abs/2202.08671


Blip glitches, a type of short-duration noise transient in the LIGO–Virgo data, are a nuisance for the binary black hole (BBH) searches. They affect the BBH search sensitivity significantly because their time-domain morphologies are very similar, and that creates difficulty in vetoing them. In this work, we construct a deep-learning neural network to efficiently distinguish BBH signals from blip glitches. We introduce sine-Gaussian projection (SGP) maps, which are projections of GW frequency-domain data snippets on a basis of sine-Gaussians defined by the quality factor and central frequency. We feed the SGP maps to our deep-learning neural network, which classifies the BBH signals and blips. Whereas the BBH signals are simulated, the blips used are taken from real data throughout our analysis. We show that our network significantly improves the identification of the BBH signals in comparison to the results obtained using traditional-$\chi^2$ and sine-Gaussian $\chi^2$. For example, our network improves the sensitivity by 75% at a false-positive rate of $10^{-2}$ for BBHs with total mass in the range $[80,140]~M_{\odot}$ and SNR in the range $[3,8]$. Also, it correctly identifies 95% of the real GW events in GWTC-3. The computation time for classification is a few minutes for thousands of SGP maps on a single core. With further optimisation in the next version of our algorithm, we expect a further reduction in the computational cost. Our proposed method can potentially improve the veto process in the LIGO–Virgo GW data analysis and conceivably support identifying GW signals in low-latency pipelines.

Read this paper on arXiv…

S. Choudhary, A. More, S. Suyamprakasam, et. al.
Fri, 18 Feb 22
3/63

Comments: 11 pages, 8 figures and 2 tables. Reviewed by LIGO Scientific Collaboration (LSC) with preprint number LIGO-P2100485

Cosmic Kite: Auto-encoding the Cosmic Microwave Background [CEA]

Posted on by

http://arxiv.org/abs/2202.05853


In this work we present the results of the study of the cosmic microwave background TT power spectrum through auto-encoders in which the latent variables are the cosmological parameters. This method was trained and calibrated using a data-set composed by 80000 power spectra from random cosmologies computed numerically with the CAMB code. Due to the specific architecture of the auto-encoder, the encoder part is a model that estimates the maximum-likelihood parameters from a given power spectrum. On the other hand, the decoder part is a model that computes the power spectrum from the cosmological parameters and can be used as a forward model in a fully Bayesian analysis. We show that the encoder is able to estimate the true cosmological parameters with a precision varying from $\approx 0.004\% $ to $\approx 0.2\% $ (depending on the cosmological parameter), while the decoder computes the power spectra with a mean percentage error of $\approx 0.0018\% $ for all the multipole range. We also demonstrate that the decoder recovers the expected trends when varying the cosmological parameters one by one, and that it does not introduce any significant bias on the estimation of cosmological parameters through a Bayesian analysis. These studies gave place to the Cosmic Kite python software that is publicly available and can be downloaded and installed from https://github.com/Martindelosrios/cosmic-kite. Although this algorithm does not improve the precision of the measurements compared with the traditional methods, it reduces significantly the computation time and represents the first attempt towards forcing the latent variables to have a physical interpretation.

Read this paper on arXiv…

M. Rios
Tue, 15 Feb 22
14/75

Comments: Accepted for its publication in MNRAS

SUPA: A Lightweight Diagnostic Simulator for Machine Learning in Particle Physics [CL]

Posted on by

http://arxiv.org/abs/2202.05012


Deep learning methods have gained popularity in high energy physics for fast modeling of particle showers in detectors. Detailed simulation frameworks such as the gold standard Geant4 are computationally intensive, and current deep generative architectures work on discretized, lower resolution versions of the detailed simulation. The development of models that work at higher spatial resolutions is currently hindered by the complexity of the full simulation data, and by the lack of simpler, more interpretable benchmarks. Our contribution is SUPA, the SUrrogate PArticle propagation simulator, an algorithm and software package for generating data by simulating simplified particle propagation, scattering and shower development in matter. The generation is extremely fast and easy to use compared to Geant4, but still exhibits the key characteristics and challenges of the detailed simulation. We support this claim experimentally by showing that performance of generative models on data from our simulator reflects the performance on a dataset generated with Geant4. The proposed simulator generates thousands of particle showers per second on a desktop machine, a speed up of up to 6 orders of magnitudes over Geant4, and stores detailed geometric information about the shower propagation. SUPA provides much greater flexibility for setting initial conditions and defining multiple benchmarks for the development of models. Moreover, interpreting particle showers as point clouds creates a connection to geometric machine learning and provides challenging and fundamentally new datasets for the field.
The code for SUPA is available at https://github.com/itsdaniele/SUPA.

Read this paper on arXiv…

A. Sinha, D. Paliotta, B. Máté, et. al.
Fri, 11 Feb 22
61/71

Comments: N/A

Assessing Planetary Complexity and Potential Agnostic Biosignatures using Epsilon Machines [EPA]

Posted on by

http://arxiv.org/abs/2202.03699


We present a new approach to exoplanet characterisation using techniques from complexity science, with potential applications to biosignature detection. This agnostic method makes use of the temporal variability of light reflected or emitted from a planet. We use a technique known as epsilon machine reconstruction to compute the statistical complexity, a measure of the minimal model size for time series data. We demonstrate that statistical complexity is an effective measure of the complexity of planetary features. Increasing levels of qualitative planetary complexity correlate with increases in statistical complexity and Shannon entropy, demonstrating that our approach can identify planets with the richest dynamics. We also compare Earth time series with Jupiter data, and find that for the three wavelengths considered, Earth’s average complexity and entropy rate are approximately 50% and 43% higher than Jupiter’s, respectively. The majority of schemes for the detection of extraterrestrial life rely upon biochemical signatures and planetary context. However, it is increasingly recognised that extraterrestrial life could be very different to life on Earth. Under the hypothesis that there is a correlation between the presence of a biosphere and observable planetary complexity, our technique offers an agnostic and quantitative method for the measurement thereof.

Read this paper on arXiv…

S. Bartlett, J. Li, L. Gu, et. al.
Wed, 9 Feb 22
16/48

Comments: N/A

Deep Learning application for stellar parameters determination: I- Constraining the hyperparameters [IMA]

Posted on by

http://arxiv.org/abs/2201.12476


Machine Learning is an efficient method for analyzing and interpreting the increasing amount of astronomical data that is available. In this study, we show, a pedagogical approach that should benefit anyone willing to experiment with Deep Learning techniques in the context of stellar parameters determination. Utilizing the Convolutional Neural Network architecture, we give a step by step overview of how to select the optimal parameters for deriving the most accurate values for the stellar parameters of stars: T$_{\rm{eff}}$, $\log g$, [X/H], and $v_e \sin i$. Synthetic spectra with random noise were used to constrain this method and to mimic the observations. We found that each stellar parameter requires a different combination of network hyperparameters and the maximum accuracy reached depends on this combination, as well as, the Signal to Noise ratio of the observations, and the architecture of the network. We also show that this technique can be applied to other spectral types in different wavelength ranges after the technique has been optimized.

Read this paper on arXiv…

M. Gebran, K. Connick, H. Farhat, et. al.
Tue, 1 Feb 22
51/73

Comments: Accepted in Open Astronomy, De Gruyter

Fully Adaptive Bayesian Algorithm for Data Analysis, FABADA [IMA]

Posted on by

http://arxiv.org/abs/2201.05145


The aim of this paper is to describe a novel non-parametric noise reduction technique from the point of view of Bayesian inference that may automatically improve the signal-to-noise ratio of one- and two-dimensional data, such as e.g. astronomical images and spectra. The algorithm iteratively evaluates possible smoothed versions of the data, the smooth models, obtaining an estimation of the underlying signal that is statistically compatible with the noisy measurements. Iterations stop based on the evidence and the $\chi^2$ statistic of the last smooth model, and we compute the expected value of the signal as a weighted average of the whole set of smooth models. In this paper, we explain the mathematical formalism and numerical implementation of the algorithm, and we evaluate its performance in terms of the peak signal to noise ratio, the structural similarity index, and the time payload, using a battery of real astronomical observations. Our Fully Adaptive Bayesian Algorithm for Data Analysis (FABADA) yields results that, without any parameter tuning, are comparable to standard image processing algorithms whose parameters have been optimized based on the true signal to be recovered, something that is impossible in a real application. State-of-the-art non-parametric methods, such as BM3D, offer slightly better performance at high signal-to-noise ratio, while our algorithm is significantly more accurate for extremely noisy data (higher than $20-40\%$ relative errors, a situation of particular interest in the field of astronomy). In this range, the standard deviation of the residuals obtained by our reconstruction may become more than an order of magnitude lower than that of the original measurements. The source code needed to reproduce all the results presented in this report, including the implementation of the method, is publicly available at https://github.com/PabloMSanAla/fabada

Read this paper on arXiv…

P. Sanchez-Alarcon and Y. Sequeiros
Fri, 14 Jan 22
44/52

Comments: 13 pages, 6 figures, submitted to IEEE Transactions on Image Processing

Orbital elements and individual component masses from joint spectroscopic and astrometric data of double-line spectroscopic binaries [SSA]

Posted on by

http://arxiv.org/abs/2201.04134


We present orbital elements, orbital parallaxes and individual component masses, for fourteen spatially resolved double-line spectroscopic binaries derived doing a simultaneous fit of their visual orbit and radial velocity curve. This was done by means of a Markov Chain Monte Carlo code developed by our group, which produces posterior distribution functions and error estimates for all the parameters. Of this sample, six systems had high quality previous studies and were included as benchmarks to test our procedures, but even in these cases we could improve the previous orbits by adding recent data from our survey of southern binaries being carried out with the HRCam and ZORRO speckle cameras at the SOAR 4.1m and Gemini South 8.1m telescopes, respectively. We also give results for eight objects that did not have a published combined orbital solution, one of which did not have a visual orbit either. We could determine mass ratios with a typical uncertainty of less than 1%, mass sums with uncertainties of about 1% and individual component masses with a formal uncertainty of $0.01 M_\odot$ in the best cases. A comparison of our orbital parallaxes with available trigonometric parallaxes from Hipparcos and Gaia eDR3, shows a good correspondence; the mean value of the differences being consistent with zero within the errors of both catalogs. We also present observational HR diagrams for our sample of binaries, which in combination with isochrones from different sources allowed us to asses their evolutionary status and also the quality of their photometry.

Read this paper on arXiv…

J. Anguita-Aguero, R. Mendez, R. Claveria, et. al.
Wed, 12 Jan 22
19/89

Comments: 26 pages, 8 figures. Accepted for publication in The Astronomical Journal

Unsupervised Machine Learning for Exploratory Data Analysis of Exoplanet Transmission Spectra [EPA]

Posted on by

http://arxiv.org/abs/2201.02696


Transit spectroscopy is a powerful tool to decode the chemical composition of the atmospheres of extrasolar planets. In this paper we focus on unsupervised techniques for analyzing spectral data from transiting exoplanets. We demonstrate methods for i) cleaning and validating the data, ii) initial exploratory data analysis based on summary statistics (estimates of location and variability), iii) exploring and quantifying the existing correlations in the data, iv) pre-processing and linearly transforming the data to its principal components, v) dimensionality reduction and manifold learning, vi) clustering and anomaly detection, vii) visualization and interpretation of the data. To illustrate the proposed unsupervised methodology, we use a well-known public benchmark data set of synthetic transit spectra. We show that there is a high degree of correlation in the spectral data, which calls for appropriate low-dimensional representations. We explore a number of different techniques for such dimensionality reduction and identify several suitable options in terms of summary statistics, principal components, etc. We uncover interesting structures in the principal component basis, namely, well-defined branches corresponding to different chemical regimes of the underlying atmospheres. We demonstrate that those branches can be successfully recovered with a K-means clustering algorithm in fully unsupervised fashion. We advocate for a three-dimensional representation of the spectroscopic data in terms of the first three principal components, in order to reveal the existing structure in the data and quickly characterize the chemical class of a planet.

Read this paper on arXiv…

K. Matchev, K. Matcheva and A. Roman
Tue, 11 Jan 22
76/95

Comments: 10 pages, 11 figures, submitted to MNRAS

Wavescan: multiresolution regression of gravitational-wave data [CL]

Posted on by

http://arxiv.org/abs/2201.01096


Identification of a transient gravitational-wave signal embedded into non-stationary noise requires the analysis of time-dependent spectral components in the resulting time series. The time-frequency distribution of the signal power can be estimated with Gabor atoms, or wavelets, localized in time and frequency by a window function. Such analysis is limited by the Heisenberg-Gabor uncertainty, which does not allow a high-resolution localization of power with individual wavelets simultaneously in time and frequency. As a result, the temporal and spectral leakage affects the time-frequency distribution, limiting the identification of sharp features in the power spectrum. This paper presents a time-frequency regression method where instead of a single window, a stack of wavelets with different windows spanning a wide range of resolutions is used to scan power at each time-frequency location. Such a wavelet scan (dubbed in the paper as wavescan) extends the conventional multiresolution analysis to capture transient signals and remove the local power variations due to the temporal and spectral leakage. A wavelet, least affected by the leakage, is selected from the stack at each time-frequency location to obtain the high-resolution localization of power. The paper presents all stages of the multiresolution wavescan regression, including the estimation of the time-varying spectrum, identification of transient signals in the time-frequency domain, and reconstruction of the corresponding time-domain waveforms. To demonstrate the performance of the method, the wavescan regression is applied to the gravitational wave data from the LIGO detectors.

Read this paper on arXiv…

S. Klimenko
Wed, 5 Jan 22
35/54

Comments: 8 pages, 8 figures

Addendum: Precision in high resolution absorption line modelling, analytic Voigt derivatives, and optimisation methods [IMA]

Posted on by

http://arxiv.org/abs/2112.14490


The parent paper to this Addendum describes the optimisation theory on which VPFIT, a non-linear least-squares program for modelling absorption spectra, is based. In that paper, we show that Voigt function derivatives can be calculated analytically using Taylor series expansions and look-up tables, for the specific case of one column density parameter for each absorption component. However, in many situations, modelling requires more complex parameterisation, such as summed column densities over a whole absorption complex, or common pattern relative ion abundances. This Addendum provides those analytic derivatives.

Read this paper on arXiv…

C. Lee, J. Webb and R. Carswell
Thu, 30 Dec 21
46/71

Comments: 4 pages, 2 figures. Submitted to MNRAS 23 Dec 2021, accepted 24 Dec 2021

Addendum: Precision in high resolution absorption line modelling, analytic Voigt derivatives, and optimisation methods [IMA]

Posted on by

http://arxiv.org/abs/2112.14490


The parent paper to this Addendum describes the optimisation theory on which VPFIT, a non-linear least-squares program for modelling absorption spectra, is based. In that paper, we show that Voigt function derivatives can be calculated analytically using Taylor series expansions and look-up tables, for the specific case of one column density parameter for each absorption component. However, in many situations, modelling requires more complex parameterisation, such as summed column densities over a whole absorption complex, or common pattern relative ion abundances. This Addendum provides those analytic derivatives.

Read this paper on arXiv…

C. Lee, J. Webb and R. Carswell
Thu, 30 Dec 21
6/71

Comments: 4 pages, 2 figures. Submitted to MNRAS 23 Dec 2021, accepted 24 Dec 2021

Analytical Modelling of Exoplanet Transit Specroscopy with Dimensional Analysis and Symbolic Regression [EPA]

Posted on by

http://arxiv.org/abs/2112.11600


The physical characteristics and atmospheric chemical composition of newly discovered exoplanets are often inferred from their transit spectra which are obtained from complex numerical models of radiative transfer. Alternatively, simple analytical expressions provide insightful physical intuition into the relevant atmospheric processes. The deep learning revolution has opened the door for deriving such analytical results directly with a computer algorithm fitting to the data. As a proof of concept, we successfully demonstrate the use of symbolic regression on synthetic data for the transit radii of generic hot Jupiter exoplanets to derive a corresponding analytical formula. As a preprocessing step, we use dimensional analysis to identify the relevant dimensionless combinations of variables and reduce the number of independent inputs, which improves the performance of the symbolic regression. The dimensional analysis also allowed us to mathematically derive and properly parametrize the most general family of degeneracies among the input atmospheric parameters which affect the characterization of an exoplanet atmosphere through transit spectroscopy.

Read this paper on arXiv…

K. Matchev, K. Matcheva and A. Roman
Thu, 23 Dec 21
36/63

Comments: Submitted to AAS Journals, 24 pages, 7 figures

Searching for Anomalies in the ZTF Catalog of Periodic Variable Stars [SSA]

Posted on by

http://arxiv.org/abs/2112.03306


Periodic variables illuminate the physical processes of stars throughout their lifetime. Wide-field surveys continue to increase our discovery rates of periodic variable stars. Automated approaches are essential to identify interesting periodic variable stars for multi-wavelength and spectroscopic follow-up. Here, we present a novel unsupervised machine learning approach to hunt for anomalous periodic variables using phase-folded light curves presented in the Zwicky Transient Facility Catalogue of Periodic Variable Stars by \citet{Chen_2020}. We use a convolutional variational autoencoder to learn a low dimensional latent representation, and we search for anomalies within this latent dimension via an isolation forest. We identify anomalies with irregular variability. Most of the top anomalies are likely highly variable Red Giants or Asymptotic Giant Branch stars concentrated in the Milky Way galactic disk; a fraction of the identified anomalies are more consistent with Young Stellar Objects. Detailed spectroscopic follow-up observations are encouraged to reveal the nature of these anomalies.

Read this paper on arXiv…

H. Chan, V. Villar, S. Cheung, et. al.
Wed, 8 Dec 21
16/77

Comments: 26 pages, 17 figures. The full version of Table 4 and Table 5 are available upon request

How to quantify fields or textures? A guide to the scattering transform [IMA]

Posted on by

http://arxiv.org/abs/2112.01288


Extracting information from stochastic fields or textures is a ubiquitous task in science, from exploratory data analysis to classification and parameter estimation. From physics to biology, it tends to be done either through a power spectrum analysis, which is often too limited, or the use of convolutional neural networks (CNNs), which require large training sets and lack interpretability. In this paper, we advocate for the use of the scattering transform (Mallat 2012), a powerful statistic which borrows mathematical ideas from CNNs but does not require any training, and is interpretable. We show that it provides a relatively compact set of summary statistics with visual interpretation and which carries most of the relevant information in a wide range of scientific applications. We present a non-technical introduction to this estimator and we argue that it can benefit data analysis, comparison to models and parameter inference in many fields of science. Interestingly, understanding the core operations of the scattering transform allows one to decipher many key aspects of the inner workings of CNNs.

Read this paper on arXiv…

S. Cheng and B. Ménard
Fri, 3 Dec 21
12/81

Comments: 18 pages, 16 figures

Probabilistic segmentation of overlapping galaxies for large cosmological surveys [IMA]

Posted on by

http://arxiv.org/abs/2111.15455


Encoder-Decoder networks such as U-Nets have been applied successfully in a wide range of computer vision tasks, especially for image segmentation of different flavours across different fields. Nevertheless, most applications lack of a satisfying quantification of the uncertainty of the prediction. Yet, a well calibrated segmentation uncertainty can be a key element for scientific applications such as precision cosmology. In this on-going work, we explore the use of the probabilistic version of the U-Net, recently proposed by Kohl et al (2018), and adapt it to automate the segmentation of galaxies for large photometric surveys. We focus especially on the probabilistic segmentation of overlapping galaxies, also known as blending. We show that, even when training with a single ground truth per input sample, the model manages to properly capture a pixel-wise uncertainty on the segmentation map. Such uncertainty can then be propagated further down the analysis of the galaxy properties. To our knowledge, this is the first time such an experiment is applied for galaxy deblending in astrophysics.

Read this paper on arXiv…

B. Hubert, B. Alexandre and H. Marc
Wed, 1 Dec 21
45/110

Comments: 7 pages, 5 figures, Accepted for the Fourth Workshop on Machine Learning and the Physical Sciences (NeurIPS 2021)

Rapid search for massive black hole binary coalescences using deep learning [IMA]

Posted on by

http://arxiv.org/abs/2111.14546


The coalescences of massive black hole binaries (MBHBs) are one of the main targets of space-based gravitational wave observatories. Such gravitational wave sources are expected to be accompanied by electromagnetic emission. Low latency time of gravitational wave searches and accurate sky localization are keys in triggering successful follow-up observations on the electromagnetic counterparts. Here we present a deep learning method for the first time to rapidly search for MBHB signals in the strain data. Our model is capable to process 1-year of data in just several seconds, identifying all MBHB coalescences with no false alarms. We test the performance of our model on the simulated data from the LISA data challenge. We demonstrate that the model shows a robust resistance for a wide range of generalization for MBHB signals. This method is supposed to be an effective approach, which combined the advances of artificial intelligence to open a new pathway for space-based gravitational wave observations.

Read this paper on arXiv…

W. Ruan, H. Wang, C. Liu, et. al.
Tue, 30 Nov 21
102/105

Comments: 8 pages, 5 figures

A Convolutional Autoencoder-Based Pipeline for Anomaly Detection and Classification of Periodic Variables [SSA]

Posted on by

http://arxiv.org/abs/2111.13828


The periodic pulsations of stars teach us about their underlying physical process. We present a convolutional autoencoder-based pipeline as an automatic approach to search for out-of-distribution anomalous periodic variables within The Zwicky Transient Facility Catalog of Periodic Variable Stars (ZTF CPVS). We use an isolation forest to rank each periodic variable by its anomaly score. Our overall most anomalous events have a unique physical origin: they are mostly highly variable and irregular evolved stars. Multiwavelength data suggest that they are most likely Red Giant or Asymptotic Giant Branch stars concentrated in the Milky Way galactic disk. Furthermore, we show how the learned latent features can be used for the classification of periodic variables through a hierarchical random forest. This novel semi-supervised approach allows astronomers to identify the most anomalous events within a given physical class, significantly increasing the potential for scientific discovery.

Read this paper on arXiv…

H. Chan, S. Cheung, V. Villar, et. al.
Tue, 30 Nov 21
105/105

Comments: 7 pages, 4 figures

Neural Network Reconstruction of Late-Time Cosmology and Null Tests [CEA]

Posted on by

http://arxiv.org/abs/2111.11462


The prospect of nonparametric reconstructions of cosmological parameters from observational data sets has been a popular topic in the literature for a number of years. This has mainly taken the form of a technique based on Gaussian processes but this approach is exposed to several foundational issues ranging from overfitting to kernel consistency problems. In this work, we explore the possibility of using artificial neural networks (ANN) to reconstruct late-time expansion and large scale structure cosmological parameters. We first show how mock data can be used to design an optimal ANN for both parameters, which we then use with real data to infer their respective redshift profiles. In addition, we also consider null tests on the reconstructed data against a concordance cosmology.

Read this paper on arXiv…

K. Dialektopoulos, J. Said, J. Mifsud, et. al.
Wed, 24 Nov 21
1/61

Comments: 28 pages, 9 figures

Empirically estimating the distribution of the loudest candidate from a gravitational-wave search [CL]

Posted on by

http://arxiv.org/abs/2111.12032


Searches for gravitational-wave signals are often based on maximizing a detection statistic over a bank of waveform templates, covering a given parameter space with a variable level of correlation. Results are often evaluated using a noise-hypothesis test, where the background is characterized by the sampling distribution of the loudest template. In the context of continuous gravitational-wave searches, properly describing said distribution is an open problem: current approaches focus on a particular detection statistic and neglect template-bank correlations. We introduce a new approach using extreme value theory to describe the distribution of the loudest template’s detection statistic in an arbitrary template bank. Our new proposal automatically generalizes to a wider class of detection statistics, including (but not limited to) line-robust statistics and transient continuous-wave signal hypotheses, and improves the estimation of the expected maximum detection statistic at a negligible computing cost. The performance of our proposal is demonstrated on simulated data as well as by applying it to different kinds of (transient) continuous-wave searches using O2 Advanced LIGO data. We release an accompanying Python software package, distromax, implementing our new developments.

Read this paper on arXiv…

R. Tenorio, L. Modafferi, D. Keitel, et. al.
Wed, 24 Nov 21
11/61

Comments: 24 pages, 23 figures, comments welcome. Package freely available in this https URL

COCOPLOT: COlor COllapsed PLOTting software : Using color to view 3D data as a 2D image [IMA]

Posted on by

http://arxiv.org/abs/2111.10786


Most modern solar observatories deliver data products formatted as 3D spatio-temporal data cubes, that contain additional, higher dimensions with spectral and/or polarimetric information. This multi-dimensional complexity presents a major challenge when browsing for features of interest in several dimensions simultaneously. We developed the COlor COllapsed PLOTting (COCOPLOT) software as a quick-look and context image software, to convey spectral profile or time evolution from all the spatial pixels ($x,y$) in a 3D [$n_x,n_y,n_\lambda$] or [$n_x,n_y,n_t$] data cube as a single image, using color. This can avoid the need to scan through many wavelengths, creating difference and composite images when searching for signals satisfying multiple criteria. Filters are generated for the red, green, and blue channels by selecting values of interest to highlight in each channel, and their weightings. These filters are combined with the data cube over the third dimension axis to produce an $n_x \times n_y \times 3$ cube displayed as one true color image. Some use cases are presented for data from the Swedish 1-m Solar Telescope (SST) and IRIS, including H$\alpha$ solar flare data, a comparison with $k$-means clustering for identifying asymmetries in the Ca II K line and off-limb coronal rain in IRIS C II slit-jaw images. These illustrate identification by color alone using COCOPLOT of locations including line wing or central enhancement, broadening, wing absorption, and sites with intermittent flows or time-persistent features. COCOPLOT is publicly available in both IDL and Python.

Read this paper on arXiv…

M. Druett, A. Pietrow, G. Vissers, et. al.
Tue, 23 Nov 21
3/84

Comments: Submitted to RASTI

Aqueous Alteration on Asteroids Simplifies Soluble Organic Matter Mixtures [EPA]

Posted on by

http://arxiv.org/abs/2111.10004


Biologically relevant abiotic extraterrestrial soluble organic matter (SOM) has been widely investigated to study the origin of life and the chemical evolution of protoplanetary disks. Synthesis of biologically relevant organics, in particular, seems to require aqueous environments in the early solar system. However, SOM in primitive meteorites includes numerous chemical species besides the biologically relevant ones, and the reaction mechanisms that comprehensively explain the complex nature of SOM are unknown. Besides, the initial reactants, which formed before asteroid accretion, were uncharacterized. We examined the mass distribution of SOM extracted from three distinct Tagish Lake meteorite fragments, which exhibit different degrees of aqueous alteration though they originated from a single asteroid. We report that mass distributions of SOM in the primordial fragments are well fit by the SchulzZimm (SZ) model for the molecular weight distribution patterns found in chain growth polymerization experiments. Also, the distribution patterns diverge further from SZ with increasing degrees of aqueous alteration. These observations imply that the complex nature of the primordial SOM (1) was established before severe alteration on the asteroid, (2) possibly existed before parent-body accretion, and (3) later became simplified on the asteroid. Therefore, aqueous reactions on asteroids are not required conditions for cultivating complex SOM. Furthermore, we found that overall H over C ratios of SOM decrease with increasing aqueous alteration, and the estimate of H loss from the SOM is 10% to 30%. Organics seem to be a significant H2 source that may have caused subsequent chemical reactions in the Tagish Lake meteorite parent body.

Read this paper on arXiv…

J. Isa, F. Orthous-Daunay, P. Beck, et. al.
Mon, 22 Nov 21
1/53

Comments: 29 pages

Nii: a Bayesian orbit retrieval code applied to differential astrometry [IMA]

Posted on by

http://arxiv.org/abs/2111.07008


Here we present an open source Python-based Bayesian orbit retrieval code (Nii) that implements an automatic parallel tempering Markov chain Monte Carlo (APT-MCMC) strategy. Nii provides a module to simulate the observations of a space-based astrometry mission in the search for exoplanets, a signal extraction process for differential astrometric measurements using multiple reference stars, and an orbital parameter retrieval framework using APT-MCMC. We further verify the orbit retrieval ability of the code through two examples corresponding to a single-planet system and a dual-planet system. In both cases, efficient convergence on the posterior probability distribution can be achieved. Although this code specifically focuses on the orbital parameter retrieval problem of differential astrometry, Nii can also be widely used in other Bayesian analysis applications.

Read this paper on arXiv…

S. Jin, X. Ding, S. Wang, et. al.
Tue, 16 Nov 21
44/97

Comments: Accepted for publication in MNRAS

Efficient estimation method for time evolution of proto-neutron star mass and radius from supernova neutrino signal [HEAP]

Posted on by

http://arxiv.org/abs/2111.05869


In this paper we present a novel method to estimate the time evolution of proto-neutron star (PNS) structure from the neutrino signal in core-collapse supernovae (CCSN). Employing recent results of multi-dimensional CCSN simulations, we delve into a relation between total emitted neutrino energy (TONE) and PNS mass/radius, and we find that they are strongly correlated with each other. We fit the relation by simple polynomial functions connecting TONE to PNS mass and radius as a function of time. By combining another fitting function representing the correlation between TONE and cumulative number of event at each neutrino observatory, PNS mass and radius can be retrieved from purely observed neutrino data. We demonstrate retrievals of PNS mass and radius from mock data of neutrino signal, and we assess the capability of our proposed method. While underlining the limitations of the method, we also discuss the importance of the joint analysis with gravitational wave signal. This would reduce uncertainties of parameter estimations in our method, and may narrow down the possible neutrino oscillation model. The proposed method is a very easy and inexpensive computation, which will be useful in real data analysis of CCSN neutrino signal.

Read this paper on arXiv…

H. Nagakura and D. Vartanyan
Fri, 12 Nov 21
3/53

Comments: Submitted to MNRAS

Calibration of Polarimetric Radar Data using the Sylvester Equation in a Pauli Basis [CL]

Posted on by

http://arxiv.org/abs/2111.04565


In this paper we develop a new approach to the calibration of polarimetric radar data based on two key ideas. The first is the use of in-scene trihedral corner reflectors not only for radiometric and geometric calibration but also to completely remove any receiver distortion components. Secondly, we then show that the remaining transmitter distortion acts as a similarity transformation of the true scattering matrix. This leads us to employ a change of base to the Pauli matrix components. We show that in this basis calibration and the effects of Faraday rotation become much simplified and for example by using reciprocity alone we can then solve for copolar channel imbalance. Finally by using an uncalibrated symmetric point target of opportunity we can estimate cross-talks and hence fully solve the calibration problem without the need for using clutter averaging or symmetry assumptions in the covariance matrix as used in many other algorithms.

Read this paper on arXiv…

S. Cloude
Tue, 9 Nov 21
23/102

Comments: 5 pages, 3 figures

A goodness-of-fit test based on a recursive product of spacings [CL]

Posted on by

http://arxiv.org/abs/2111.02252


We introduce a new statistical test based on the observed spacings of ordered data. The statistic is sensitive to detect non-uniformity in random samples, or short-lived features in event time series. Under some conditions, this new test can outperform existing ones, such as the well known Kolmogorov-Smirnov or Anderson-Darling tests, in particular when the number of samples is small and differences occur over a small quantile of the null hypothesis distribution. A detailed description of the test statistic is provided including an illustration and examples, together with a parameterization of its distribution based on simulation.

Read this paper on arXiv…

P. Eller and L. Shtembari
Thu, 4 Nov 21
22/73

Comments: N/A

Predicting resolved galaxy properties from photometric images using convolutional neural networks [GA]

Posted on by

http://arxiv.org/abs/2111.01154


Multi-band images of galaxies reveal a huge amount of information about their morphology and structure. However, inferring properties of the underlying stellar populations such as age, metallicity or kinematics from those images is notoriously difficult. Traditionally such information is best extracted from expensive spectroscopic observations. Here we present the $Painting\, IntrinsiC\, Attributes\, onto\, SDSS\, Objects$ (PICASSSO) project and test the information content of photometric multi-band images of galaxies. We train a convolutional neural network on 27,558 galaxy image pairs to establish a connection between broad-band images and the underlying physical stellar and gaseous galaxy property maps. We test our machine learning (ML) algorithm with SDSS $ugriz$ mock images for which uncertainties and systematics are exactly known. We show that multi-band galaxy images contain enough information to reconstruct 2d maps of stellar mass, metallicity, age and gas mass, metallicity as well as star formation rate. We recover the true stellar properties on a pixel by pixel basis with only little scatter, $\lesssim20\%$ compared to $\sim50\%$ statistical uncertainty from traditional mass-to-light-ratio based methods. We further test for any systematics of our algorithm with image resolution, training sample size or wavelength coverage. We find that galaxy morphology alone constrains stellar properties to better than $\sim20\%$ thus highlighting the benefits of including morphology into the parameter estimation. The machine learning approach can predict maps of high resolution, only limited by the resolution of the input bands, thus achieving higher resolution than IFU observations. The network architecture and all code is publicly available on GitHub.

Read this paper on arXiv…

T. Buck and S. Wolf
Wed, 3 Nov 21
1/106

Comments: 14 pages main text, 9 figures, 3 pages appendix with 5 figures, submitted to MNRAS

Reproducing size distributions of swarms of barchan dunes on Mars and Earth using a mean-field model [CL]

Posted on by

http://arxiv.org/abs/2110.15850


We apply a mean-field model of interactions between migrating barchan dunes, the CAFE model, which includes calving, aggregation, fragmentation, and mass-exchange, yielding a steady-state size distribution that can be resolved for different choices of interaction parameters. The CAFE model is applied to empirically measured distributions of dune sizes in two barchan swarms on Mars, three swarms in Morocco, and one in Mauritania, each containing ~1000 bedforms, comparing the observed size distributions to the steady-states of the CAFE model. We find that the distributions in the Martian swarm are very similar to the swarm measured in Mauritania, suggesting that the two very different planetary environments however share similar dune interaction dynamics. Optimisation of the model parameters of three specific configurations of the CAFE model shows that the fit of the theoretical steady-state is often superior to the typically assumed log-normal. In all cases, the optimised parameters indicate that mass-exchange is the most frequent type of interaction. Calving is found to occur rarely in most of the swarms, with a highest rate of only 9\% of events, showing that interactions between multiple dunes rather than spontaneous calving are the driver of barchan size distributions. Finally, the implementation of interaction parameters derived from 3D simulations of dune-pair collisions indicates that sand flux between dunes is more important in producing the size distributions of the Moroccan swarms than of those in Mauritania and on Mars.

Read this paper on arXiv…

D. Robson, A. Annibale and A. Baas
Mon, 1 Nov 21
21/58

Comments: 30 Pages, 10 figures, Submitted to Physica A: Statistical Mechanics and its Applications

A machine learning algorithm for direct detection of axion-like particle domain walls [IMA]

Posted on by

http://arxiv.org/abs/2110.00139


The Global Network of Optical Magnetometers for Exotic physics searches (GNOME) conducts an experimental search for certain forms of dark matter based on their spatiotemporal signatures imprinted on a global array of synchronized atomic magnetometers. The experiment described here looks for a gradient coupling of axion-like particles (ALPs) with proton spins as a signature of locally dense dark matter objects such as domain walls. In this work, stochastic optimization with machine learning is proposed for use in a search for ALP domain walls based on GNOME data. The validity and reliability of this method were verified using binary classification. The projected sensitivity of this new analysis method for ALP domain-wall crossing events is presented.

Read this paper on arXiv…

D. Kim, D. Kimball, H. Masia-Roig, et. al.
Mon, 4 Oct 21
60/76

Comments: N/A

Life, the universe and the hidden meaning of everything [CL]

Posted on by

http://arxiv.org/abs/2109.10241


It is hard to look at the universe and not wonder about the meaning, of, well, everything. A natural question is whether what we see is a sign of intelligent design. The antithesis of design would be a random universe or, assuming laws of physics, one whose fundamental physical parameters were randomly selected, but conditioned on life (ourselves) being here to observe it. In unpublished work, the British physicist Dennis Sciama argued that such a randomly selected universe would display a statistical signature. He concluded that a random universe would almost certainly have parameters only just allowing for the possibility of life. Here we consider whether this signature is definitive. We find that with plausible additional assumptions Sciama’s signature would appear to reverse: Were our universe random, it could give the false impression of being intelligently designed, with the fundamental constants appearing to be fine-tuned to a strong probability for life to emerge and be maintained.

Read this paper on arXiv…

Z. Wang and S. Braunstein
Wed, 22 Sep 21
57/57

Comments: 6 pages, 2 figures. Request for comments

Bias of the Hubble constant value caused by errors in galactic distance indicators [CEA]

Posted on by

http://arxiv.org/abs/2109.09645


The bias in the determination of the Hubble parameter and the Hubble constant in the modern Universe is discussed. It could appear due to statistical processing of data on galaxies redshifts and estimated distances based on some statistical relations with limited accuracy. This causes a number of effects leading to either underestimation or overestimation of the Hubble parameter when using any methods of statistical processing, primarily the least squares method (LSM). The value of the Hubble constant is underestimated when processing a whole sample; when the sample is constrained by distance, especially when constrained from above, it is significantly overestimated due to data selection. The bias significantly exceeds the values of the error the Hubble constant calculated by the LSM formulae.
These effects are demonstrated both analytically and using Monte Carlo simulations, which introduce deviations in both velocities and estimated distances to the original dataset described by the Hubble law. The characteristics of the deviations are similar to real observations. Errors in estimated distances are up to 20%. They lead to the fact that when processing the same mock sample using LSM, it is possible to obtain an estimate of the Hubble constant from 96% of the true value when processing the entire sample to 110% when processing the subsample with distances limited from above.
The impact of these effects can lead to a bias in the Hubble constant obtained from real data and an overestimation of the accuracy of determining this value. This may call into question the accuracy of determining the Hubble constant and significantly reduce the tension between the values obtained from the observations in the early and modern Universe, which were actively discussed during the last year.

Read this paper on arXiv…

S. S.L.Parnovsky
Tue, 21 Sep 21
79/85

Comments: 18 pages, 8 figures. Accepted for publication at Ukr. J. Phys

The Dynamic Time Warping as a means to assess solar wind time series [SSA]

Posted on by

http://arxiv.org/abs/2109.07873


During the last decades there is a continuing international endeavor in developing realistic space weather prediction tools aiming to forecast the conditions on the Sun and in the interplanetary environment. These efforts have led to the need of developing appropriate metrics in order to assess the performance of those tools. Metrics are necessary for validating models, comparing different models and monitoring adjustments or improvements of a certain model over time. In this work, we introduce the Dynamic Time Warping (DTW) as an alternative way to validate models and, in particular, to quantify differences between observed and synthetic (modeled) time series for space weather purposes. We present the advantages and drawbacks of this method as well as applications on WIND observations and EUHFORIA modeled output at L1. We show that DTW is a useful tool that permits the evaluation of both the fast and slow solar wind. Its distinctive characteristic is that it warps sequences in time, aiming to align them with the minimum cost by using dynamic programming. It can be applied in two different ways for the evaluation of modeled solar wind time series. The first way calculates the so-called sequence similarity factor (SSF), a number that provides a quantification of how good the forecast is, compared to a best and a worst case prediction scenarios. The second way quantifies the time and amplitude differences between the points that are best matched between the two sequences. As a result, it can serve as a hybrid metric between continuous measurements (such as, e.g., the correlation coefficient) and point-by-point comparisons. We conclude that DTW is a promising technique for the assessment of solar wind profiles offering functions that other metrics do not, so that it can give at once the most complete evaluation profile of a model.

Read this paper on arXiv…

E. Samara, B. Laperre, R. Kieokaew, et. al.
Fri, 17 Sep 21
25/67

Comments: N/A

Eigenspectra of solar active region long-period oscillations [SSA]

Posted on by

http://arxiv.org/abs/2109.04189


We studied the low-frequency $\lesssim 0.5\;$h$^{-1}$ (long-period $\gtrsim 2\;$h) oscillations of active regions (ARs). The investigation is based on an analysis of a time series built from Solar Dynamics Observatory/Helioseismic and Magnetic Imager (SDO/HMI) photospheric magnetograms and comprises case studies of several types of AR structures. The main goals are to investigate whether ARs can be engaged in long-period oscillations as unified oscillatory entities and, if so, to determine the spectral pattern of such oscillations. Time series of characteristic parameters of the ARs, such as, the total area, total unsigned radial magnetic flux, and tilt angle, were measured and recorded using the image moment method. The power spectra were built out of Gaussian-apodised and zero-padded datasets. There are long-period oscillations ranging from 2 to 20 h, similarly to the characteristic lifetimes of super-granulation, determined from the datasets of the AR total area and radial magnetic flux, respectively. However, no periodicity in tilt angle data was found. Whatever nature these oscillations have, they must be energetically supported by convective motions beneath the solar surface. The possible interpretations can be related to different types of magnetohydrodynamic (MHD) oscillations of the multi-scale structure of the AR magnetic field, which is probably linked with the characteristic turnover timescales of the super-granulation cells. The presence of oscillations in the radial magnetic flux data may be connected to periodic flux emergence or cancellation processes.

Read this paper on arXiv…

G. Dumbadze, B. Shergelashvili, S. Poedts, et. al.
Fri, 10 Sep 21
40/59

Comments: 11 Pages, 5 figures

Bayesian Estimation of the D(p,$γ$)$^3$He Thermonuclear Reaction Rate [CEA]

Posted on by

http://arxiv.org/abs/2109.00049


Big bang nucleosynthesis (BBN) is the standard model theory for the production of the light nuclides during the early stages of the universe, taking place for a period of about 20 minutes after the big bang. Deuterium production, in particular, is highly sensitive to the primordial baryon density and the number of neutrino species, and its abundance serves as a sensitive test for the conditions in the early universe. The comparison of observed deuterium abundances with predicted ones requires reliable knowledge of the relevant thermonuclear reaction rates, and their corresponding uncertainties. Recent observations reported the primordial deuterium abundance with percent accuracy, but some theoretical predictions based on BBN are at tension with the measured values because of uncertainties in the cross section of the deuterium-burning reactions. In this work, we analyze the S-factor of the D(p,$\gamma$)$^3$He reaction using a hierarchical Bayesian model. We take into account the results of eleven experiments, spanning the period of 1955–2021; more than any other study. We also present results for two different fitting functions, a two-parameter function based on microscopic nuclear theory and a four-parameter polynomial. Our recommended reaction rates have a 2.2\% uncertainty at $0.8$~GK, which is the temperature most important for deuterium BBN. Differences between our rates and previous results are discussed.

Read this paper on arXiv…

J. Moscoso, R. Souza, A. Coc, et. al.
Thu, 2 Sep 21
33/59

Comments: N/A

Hardware-accelerated Inference for Real-Time Gravitational-Wave Astronomy [CL]

Posted on by

http://arxiv.org/abs/2108.12430


The field of transient astronomy has seen a revolution with the first gravitational-wave detections and the arrival of multi-messenger observations they enabled. Transformed by the first detection of binary black hole and binary neutron star mergers, computational demands in gravitational-wave astronomy are expected to grow by at least a factor of two over the next five years as the global network of kilometer-scale interferometers are brought to design sensitivity. With the increase in detector sensitivity, real-time delivery of gravitational-wave alerts will become increasingly important as an enabler of multi-messenger followup. In this work, we report a novel implementation and deployment of deep learning inference for real-time gravitational-wave data denoising and astrophysical source identification. This is accomplished using a generic Inference-as-a-Service model that is capable of adapting to the future needs of gravitational-wave data analysis. Our implementation allows seamless incorporation of hardware accelerators and also enables the use of commercial or private (dedicated) as-a-service computing. Based on our results, we propose a paradigm shift in low-latency and offline computing in gravitational-wave astronomy. Such a shift can address key challenges in peak-usage, scalability and reliability, and provide a data analysis platform particularly optimized for deep learning applications. The achieved sub-millisecond scale latency will also be relevant for any machine learning-based real-time control systems that may be invoked in the operation of near-future and next generation ground-based laser interferometers, as well as the front-end collection, distribution and processing of data from such instruments.

Read this paper on arXiv…

A. Gunny, D. Rankin, J. Krupa, et. al.
Tue, 31 Aug 21
60/73

Comments: 21 pages, 14 figures

Precision in high resolution absorption line modelling, analytic Voigt derivatives, and optimisation methods [IMA]

Posted on by

http://arxiv.org/abs/2108.11218


This paper describes the optimisation theory on which VPFIT, a non-linear least-squares program for modelling absorption spectra, is based. Particular attention is paid to precision. Voigt function derivatives have previously been calculated using numerical finite difference approximations. We show how these can instead be computed analytically using Taylor series expansions and look-up tables. We introduce a new optimisation method for an efficient descent path to the best-fit, combining the principles used in both the Gauss-Newton and Levenberg-Marquardt algorithms. A simple practical fix for ill-conditioning is described, a common problem when modelling quasar absorption systems. We also summarise how unbiased modelling depends on using an appropriate information criterion to guard against over- or under-fitting.
The methods and the new implementations introduced in this paper are aimed at optimal usage of future data from facilities such as ESPRESSO/VLT and HIRES/ELT, particularly for the most demanding applications such as searches for spacetime variations in fundamental constants and attempts to detect cosmological redshift drift.

Read this paper on arXiv…

J. Webb, R. Carswell and C. Lee
Thu, 26 Aug 21
29/52

Comments: 15 pages, 7 figures, submitted to MNRAS

Self-Calibrating the Look-Elsewhere Effect: Fast Evaluation of the Statistical Significance Using Peak Heights [IMA]

Posted on by

http://arxiv.org/abs/2108.06333


In experiments where one searches a large parameter space for an anomaly, one often finds many spurious noise-induced peaks in the likelihood. This is known as the look-elsewhere effect, and must be corrected for when performing statistical analysis. This paper introduces a method to calibrate the false alarm probability (FAP), or $p$-value, for a given dataset by considering the heights of the highest peaks in the likelihood. In the simplest form of self-calibration, the look-elsewhere-corrected $\chi^2$ of a physical peak is approximated by the $\chi^2$ of the peak minus the $\chi^2$ of the highest noise-induced peak. Generalizing this concept to consider lower peaks provides a fast method to quantify the statistical significance with improved accuracy. In contrast to alternative methods, this approach has negligible computational cost as peaks in the likelihood are a byproduct of every peak-search analysis. We apply to examples from astronomy, including planet detection, periodograms, and cosmology.

Read this paper on arXiv…

A. Bayer, U. Seljak and J. Robnik
Mon, 16 Aug 21
1/34

Comments: 12 pages, 7 figures

Angular clustering properties of the DESI QSO target selection using DR9 Legacy Imaging Surveys [CEA]

Posted on by

http://arxiv.org/abs/2108.03640


The quasar target selection for the upcoming survey of the Dark Energy Spectroscopic Instrument (DESI) will be fixed for the next five years. The aim of this work is to validate the quasar selection by studying the impact of imaging systematics as well as stellar and galactic contaminants, and to develop a procedure to mitigate them. Density fluctuations of quasar targets are found to be related to photometric properties such as seeing and depth of the Data Release 9 of the DESI Legacy Imaging Surveys. To model this complex relation, we explore machine learning algorithms (Random Forest and Multi-Layer Perceptron) as an alternative to the standard linear regression. Splitting the footprint of the Legacy Imaging Surveys into three regions according to photometric properties, we perform an independent analysis in each region, validating our method using eBOSS EZ-mocks. The mitigation procedure is tested by comparing the angular correlation of the corrected target selection on each photometric region to the angular correlation function obtained using quasars from the Sloan Digital Sky Survey (SDSS)Data Release 16. With our procedure, we recover a similar level of correlation between DESI quasar targets and SDSS quasars in two thirds of the total footprint and we show that the excess of correlation in the remaining area is due to a stellar contamination which should be removed with DESI spectroscopic data. We derive the Limber parameters in our three imaging regions and compare them to previous measurements from SDSS and the 2dF QSO Redshift Survey.

Read this paper on arXiv…

E. Chaussidon, C. Yèche, N. Palanque-Delabrouille, et. al.
Tue, 10 Aug 21
28/84

Comments: 20 pages, 23 figures

A Bayesian inference and model selection algorithm with an optimisation scheme to infer the model noise power [IMA]

Posted on by

http://arxiv.org/abs/2108.02894


Model fitting is possibly the most extended problem in science. Classical approaches include the use of least-squares fitting procedures and maximum likelihood methods to estimate the value of the parameters in the model. However, in recent years, Bayesian inference tools have gained traction. Usually, Markov chain Monte Carlo methods are applied to inference problems, but they present some disadvantages, particularly when comparing different models fitted to the same dataset. Other Bayesian methods can deal with this issue in a natural and effective way. We have implemented an importance sampling algorithm adapted to Bayesian inference problems in which the power of the noise in the observations is not known a priori. The main advantage of importance sampling is that the model evidence can be derived directly from the so-called importance weights — while MCMC methods demand considerable postprocessing. The use of our adaptive target, adaptive importance sampling (ATAIS) method is shown by inferring, on the one hand, the parameters of a simulated flaring event which includes a damped oscillation {and, on the other hand, real data from the Kepler mission. ATAIS includes a novel automatic adaptation of the target distribution. It automatically estimates the variance of the noise in the model. ATAIS admits parallelisation, which decreases the computational run-times notably. We compare our method against a nested sampling method within a model selection problem.

Read this paper on arXiv…

J. Lopez-Santiago, L. Martino, J. Miguez, et. al.
Mon, 9 Aug 21
35/51

Comments: This article has been accepted for publication in MNRAS, published by Oxford University Press on behalf of the Royal Astronomical Society

Arby $-$ Fast data-driven surrogates [CL]

Posted on by

http://arxiv.org/abs/2108.01305


The availability of fast to evaluate and reliable predictive models is highly relevant in multi-query scenarios where evaluating some quantities in real, or near-real-time becomes crucial. As a result, reduced-order modelling techniques have gained traction in many areas in recent years. We introduce Arby, an entirely data-driven Python package for building reduced order or surrogate models. In contrast to standard approaches, which involve solving partial differential equations, Arby is entirely data-driven. The package encompasses several tools for building and interacting with surrogate models in a user-friendly manner. Furthermore, fast model evaluations are possible at a minimum computational cost using the surrogate model. The package implements the Reduced Basis approach and the Empirical Interpolation Method along a classic regression stage for surrogate modelling. We illustrate the simplicity in using Arby to build surrogates through a simple toy model: a damped pendulum. Then, for a real case scenario, we use Arby to describe CMB temperature anisotropies power spectra. On this multi-dimensional setting, we find that out from an initial set of $80,000$ power spectra solutions with $3,000$ multipole indices each, could be well described at a given tolerance error, using just a subset of $84$ solutions.

Read this paper on arXiv…

A. Villanueva, M. Beroiz, J. Cabral, et. al.
Wed, 4 Aug 21
49/66

Comments: 10 pages, 8 figures

An efficient hit finding algorithm for Baikal-GVD muon reconstruction [IMA]

Posted on by

http://arxiv.org/abs/2108.00208


The Baikal-GVD is a large scale neutrino telescope being constructed in Lake Baikal. The majority of signal detected by the telescope are noise hits, caused primarily by the luminescence of the Baikal water. Separating noise hits from the hits produced by Cherenkov light emitted from the muon track is a challenging part of the muon event reconstruction. We present an algorithm that utilizes a known directional hit causality criterion to contruct a graph of hits and then use a clique-based technique to select the subset of signal hits.The algorithm was tested on realistic detector Monte-Carlo simulation for a wide range of muon energies and has proved to select a pure sample of PMT hits from Cherenkov photons while retaining above 90\% of original signal.

Read this paper on arXiv…

V. Allakhverdyan, A. Avrorin, A. Avrorin, et. al.
Tue, 3 Aug 21
30/90

Comments: Presented at the 37th International Cosmic Ray Conference (ICRC 2021)

Source-Agnostic Gravitational-Wave Detection with Recurrent Autoencoders [CL]

Posted on by

http://arxiv.org/abs/2107.12698


We present an application of anomaly detection techniques based on deep recurrent autoencoders to the problem of detecting gravitational wave signals in laser interferometers. Trained on noise data, this class of algorithms could detect signals using an unsupervised strategy, i.e., without targeting a specific kind of source. We develop a custom architecture to analyze the data from two interferometers. We compare the obtained performance to that obtained with other autoencoder architectures and with a convolutional classifier. The unsupervised nature of the proposed strategy comes with a cost in terms of accuracy, when compared to more traditional supervised techniques. On the other hand, there is a qualitative gain in generalizing the experimental sensitivity beyond the ensemble of pre-computed signal templates. The recurrent autoencoder outperforms other autoencoders based on different architectures. The class of recurrent autoencoders presented in this paper could complement the search strategy employed for gravitational wave detection and extend the reach of the ongoing detection campaigns.

Read this paper on arXiv…

E. Moreno, J. Vlimant, M. Spiropulu, et. al.
Wed, 28 Jul 21
17/68

Comments: 16 pages, 6 figures

Everything you always wanted to know about matched filters (but were afraid to ask) [IMA]

Posted on by

http://arxiv.org/abs/2107.09378


In this paper we review the application of the matched filter (MF) technique and its application to detect weak, deterministic, smooth signals in a stationary, random, Gaussian noise. This is particular suitable in astronomy to detect emission lines in spectra and point-sources in two-dimensional maps. A detailed theoretical development is already available in many books (e.g. Kay 1998; Poor 1994; McNicol 2005; Hippenstiel 2002; Macmillan & Creelma 2005; Wickens 2002; Barkat 2005; Tuzlukov 2001; Levy 2008). Our aim is to examine some practical issues that are typically ignored in textbooks or even in specialized literature as, for example, the effects of the discretization of the signals and the non-Gaussian nature of the noise. To this goal we present each item in the form of answers to specific questions. The relative mathematics and its demonstration are kept to a bare simplest minimum, in the hope of a better understanding of the real performances of the MF in practical applications. For the ease of formalism, arguments will be developed for one-dimensional signals. The extension to the two-dimensional signals is trivial and will be highlighted in dedicated sections.

Read this paper on arXiv…

R. Vio and P. Andreani
Wed, 21 Jul 21
29/83

Comments: 23 pages, 41 figures, Review paper about matched filters and their applications

The Completed SDSS-IV extended Baryon Oscillation Spectroscopic Survey: The Damped Lyman-$α$ systems Catalog [CEA]

Posted on by

http://arxiv.org/abs/2107.09612


We present the characteristics of the Damped Lyman-$\alpha$ (DLA) systems found in the data release DR16 of the extended Baryon Oscillation Spectroscopic Survey (eBOSS) of the Sloan Digital Sky Survey (SDSS). DLAs were identified using the convolutional neural network (CNN) of~\cite{Parks2018}. A total of 117,458 absorber candidates were found with $2 \leq \zdla \leq 5.5$ and $19.7 \leq \lognhi \leq 22$, including 57,136 DLA candidates with $\lognhi \geq 20.3$. Mock quasar spectra were used to estimate DLA detection efficiency and the purity of the resulting catalog. Restricting the quasar sample to bright forests, i.e. those with mean forest fluxes $\meanflux>2\times\fluxunit$, the completeness and purity are greater than 90\% for DLAs with column densities in the range $20.1\leq \lognhi \leq 22$.

Read this paper on arXiv…

S. Chabanier, T. Etourneau, J. Goff, et. al.
Wed, 21 Jul 21
47/83

Comments: Submitted to ApJS

Gwadaptive_scattering: an automated pipeline for scattered light noise characterization [IMA]

Posted on by

http://arxiv.org/abs/2107.07565


Scattered light noise affects the sensitivity of gravitational waves detectors. The characterization of such noise is needed to mitigate it. The time-varying filter empirical mode decomposition algorithm is suitable for identifying signals with time-dependent frequency such as scattered light noise (or scattering). We present a fully automated pipeline based on the pytvfemd library, a python implementation of the tvf-EMD algorithm, to identify objects inducing scattering in the gravitational-wave channel with their motion. The pipeline application to LIGO Livingston O3 data shows that most scattering noise is due to the penultimate mass at the end of the X-arm of the detector (EXPUM) and with a motion in the micro-seismic frequency range.

Read this paper on arXiv…

S. Bianchi, A. Longo, G. Valdes, et. al.
Mon, 19 Jul 21
2/70

Comments: 14 pages, 8 figures

Improved Measurements of the Sun's Meridional Flow and Torsional Oscillation from Correlation tracking on MDI \& HMI magnetograms [SSA]

Posted on by

http://arxiv.org/abs/2107.07731


The Sun’s axisymmetric flows, differential rotation and meridional flow, govern the dynamics of the solar magnetic cycle and variety of methods are used to measure these flows, each with its own strengths and weaknesses. Flow measurements based on cross-correlating images of the surface magnetic field have been made since the 1970s which require advanced numerical techniques that are capable of detecting movements of less than the pixel size in images of the Sun. We have identified several systematic errors in addition to the center-to-limb effect that influence previous measurements of these flows and propose numerical techniques that can minimize these errors by utilizing measurements of displacements at several time-lags. Our analysis of line-of-sight magnetograms from the {\em Michelson Doppler Imager} (MDI) on the ESA/NASA {\em Solar and Heliospheric Observatory} (SOHO) and {\em Helioseismic and Magnetic Imager} (HMI) on the NASA {\em Solar Dynamics Observatory} (SDO) shows long-term variations in the meridional flow and differential rotation over two sunspot cycles from 1996 to 2020. These improved measurements can serve as vital inputs for solar dynamo and surface flux transport simulations.

Read this paper on arXiv…

S. Mahajan, D. Hathaway, A. Muñoz-Jaramillo, et. al.
Mon, 19 Jul 21
36/70

Comments: 16 pages, 9 figures, 2 tables

Shared Data and Algorithms for Deep Learning in Fundamental Physics [CL]

Posted on by

http://arxiv.org/abs/2107.00656


We introduce a collection of datasets from fundamental physics research — including particle physics, astroparticle physics, and hadron- and nuclear physics — for supervised machine learning studies. These datasets, containing hadronic top quarks, cosmic-ray induced air showers, phase transitions in hadronic matter, and generator-level histories, are made public to simplify future work on cross-disciplinary machine learning and transfer learning in fundamental physics. Based on these data, we present a simple yet flexible graph-based neural network architecture that can easily be applied to a wide range of supervised learning tasks in these domains. We show that our approach reaches performance close to state-of-the-art dedicated methods on all datasets. To simplify adaptation for various problems, we provide easy-to-follow instructions on how graph-based representations of data structures, relevant for fundamental physics, can be constructed and provide code implementations for several of them. Implementations are also provided for our proposed method and all reference algorithms.

Read this paper on arXiv…

L. Benato, E. Buhmann, M. Erdmann, et. al.
Mon, 5 Jul 21
18/52

Comments: 13 pages, 5 figures, 5 tables

Mapping the Likelihood of GW190521 with Diverse Mass and Spin Priors [HEAP]

Posted on by

http://arxiv.org/abs/2106.13821


We map the likelihood of GW190521, the heaviest detected binary black hole (BBH) merger, by sampling under different mass and spin priors designed to be uninformative. We find that a source-frame total mass of $\sim$$150 M_{\odot}$ is consistently supported, but posteriors in mass ratio and spin depend critically on the choice of priors. We confirm that the likelihood has a multi-modal structure with peaks in regions of mass ratio representing very different astrophysical scenarios. The unequal-mass region ($m_2 / m_1 < 0.3$) has an average likelihood $\sim$$e^6$ times larger than the equal-mass region ($m_2 / m_1 > 0.3$) and a maximum likelihood $\sim$$e^2$ larger. Using ensembles of samples across priors, we examine the implications of qualitatively different BBH sources that fit the data. We find that the equal-mass solution has poorly constrained spins and at least one black hole mass that is difficult to form via stellar collapse due to pair instability. The unequal-mass solution can avoid this mass gap entirely but requires a negative effective spin and a precessing primary. Either of these scenarios is more easily produced by dynamical formation channels than field binary co-evolution. The sensitive comoving volume-time of the mass gap solution is $\mathcal{O}(10)$ times larger than the gap-avoiding solution. After accounting for this distance effect, the likelihood still reverses the advantage to favor the gap-avoiding scenario by a factor of $\mathcal{O}(100)$ before considering mass and spin priors. Posteriors are easily driven away from this high-likelihood region by common prior choices meant to be uninformative, making GW190521 parameter inference sensitive to the assumed mass and spin distributions of mergers in the source’s astrophysical channel. This may be a generic issue for similarly heavy events given current detector sensitivity and waveform degeneracies.

Read this paper on arXiv…

S. Olsen, J. Roulet, H. Chia, et. al.
Tue, 29 Jun 21
58/101

Comments: 13 pages, 10 figures

Primordial non-Gaussianity from the Completed SDSS-IV extended Baryon Oscillation Spectroscopic Survey I: Catalogue Preparation and Systematic Mitigation [CEA]

Posted on by

http://arxiv.org/abs/2106.13724


We investigate the large-scale clustering of the final spectroscopic sample of quasars from the recently completed extended Baryon Oscillation Spectroscopic Survey (eBOSS). The sample contains $343708$ objects in the redshift range $0.8<z<2.2$ and $72667$ objects with redshifts $2.2<z<3.5$, covering an effective area of $4699~{\rm deg}^{2}$. We develop a neural network-based approach to mitigate spurious fluctuations in the density field caused by spatial variations in the quality of the imaging data used to select targets for follow-up spectroscopy. Simulations are used with the same angular and radial distributions as the real data to estimate covariance matrices, perform error analyses, and assess residual systematic uncertainties. We measure the mean density contrast and cross-correlations of the eBOSS quasars against maps of potential sources of imaging systematics to address algorithm effectiveness, finding that the neural network-based approach outperforms standard linear regression. Stellar density is one of the most important sources of spurious fluctuations, and a new template constructed using data from the Gaia spacecraft provides the best match to the observed quasar clustering. The end-product from this work is a new value-added quasar catalogue with the improved weights to correct for nonlinear imaging systematic effects, which will be made public. Our quasar catalogue is used to measure the local-type primordial non-Gaussianity in our companion paper, Mueller et al. in preparation.

Read this paper on arXiv…

M. Rezaie, A. Ross, H. Seo, et. al.
Mon, 28 Jun 21
12/51

Comments: 17 pages, 13 figures, 2 tables. Accepted for publication in MNRAS. For the associated code and value-added catalogs see this https URL and this https URL

On a new statistical technique for the real-time recognition of ultra-low multiplicity astrophysical neutrino burst [IMA]

Posted on by

http://arxiv.org/abs/2106.12345


The real-time recognition of neutrino signals from astrophysical objects with very-low false alarm rate and short-latency, is crucial to perform multi-messenger detection, especially in the case of distant core-collapse supernovae accessible with the next generation of large-scale neutrino telescopes. The current time-based selection algorithms implemented in operating online monitors depend mainly on the number of events (multiplicity) detected in a fixed time window, under the hypothesis of Poisson-distributed background. However, these methods are not capable of exploiting the time profile discrepancies between the expected supernova neutrino burst and the stationary background.
In this paper we propose a new general and flexible technique (beta filter method) which provides specific decision boundaries on the cluster multiplicity-duration plane, guaranteeing the desired false alarm rate in an analytical way. The performance is evaluated using the injection of a general purpose SN-like signal on top of realistic background rates in current detectors. An absolute gain in efficiency of up to $\sim 80\%$ is achieved compared with the standard techniques, and a new ultra-low multiplicity region is unveiled.

Read this paper on arXiv…

M. Mattiazzi, M. Lamoureux and G. Collazuol
Thu, 24 Jun 21
44/54

Comments: 7 pages, 7 figures

Efficient Computation of $N$-point Correlation Functions in $D$ Dimensions [IMA]

Posted on by

http://arxiv.org/abs/2106.10278


We present efficient algorithms for computing the $N$-point correlation functions (NPCFs) of random fields in arbitrary $D$-dimensional homogeneous and isotropic spaces. Such statistics appear throughout the physical sciences, and provide a natural tool to describe a range of stochastic processes. Typically, NPCF estimators have $\mathcal{O}(n^N)$ complexity (for a data set containing $n$ particles); their application is thus computationally infeasible unless $N$ is small. By projecting onto a suitably-defined angular basis, we show that the estimators can be written in separable form, with complexity $\mathcal{O}(n^2)$, or $\mathcal{O}(n_{\rm g}\log n_{\rm g})$ if evaluated using a Fast Fourier Transform on a grid of size $n_{\rm g}$. Our decomposition is built upon the $D$-dimensional hyperspherical harmonics; these form a complete basis on the $(D-1)$-sphere and are intrinsically related to angular momentum operators. Concatenation of $(N-1)$ such harmonics gives states of definite combined angular momentum, forming a natural separable basis for the NPCF. In particular, isotropic correlation functions require only states with zero combined angular momentum. We provide explicit expressions for the NPCF estimators as applied to both discrete and gridded data, and discuss a number of applications within cosmology and fluid dynamics. The efficiency of such estimators will allow higher-order correlators to become a standard tool in the analysis of random fields.

Read this paper on arXiv…

O. Philcox and Z. Slepian
Tue, 22 Jun 21
62/71

Comments: 12 pages, 2 figures, submitted to PNAS. Comments welcome!

Fasano-Franceschini Test: an Implementation of a 2-Dimensional Kolmogorov-Smirnov test in R [CL]

Posted on by

http://arxiv.org/abs/2106.10539


The univariate Kolmogorov-Smirnov (KS) test is a non-parametric statistical test designed to assess whether a set of data is consistent with a given probability distribution (or, in the two-sample case, whether the two samples come from the same underlying distribution). The versatility of the KS test has made it a cornerstone of statistical analysis and is commonly used across the scientific disciplines. However, the test proposed by Kolmogorov and Smirnov does not naturally extend to multidimensional distributions. Here, we present the fasano.franceschini.test package, an R implementation of the 2-D KS two-sample test as defined by Fasano and Franceschini (Fasano and Franceschini 1987). The fasano.franceschini.test package provides three improvements over the current 2-D KS test on the Comprehensive R Archive Network (CRAN): (i) the Fasano and Franceschini test has been shown to run in $O(n^2)$ versus the Peacock implementation which runs in $O(n^3)$; (ii) the package implements a procedure for handling ties in the data; and (iii) the package implements a parallelized bootstrapping procedure for improved significance testing. Ultimately, the fasano.franceschini.test package presents a robust statistical test for analyzing random samples defined in 2-dimensions.

Read this paper on arXiv…

E. Ness-Cohn and R. Braun
Tue, 22 Jun 21
71/71

Comments: 8 pages, 4 figures

Maximum Entropy Spectral Analysis: a case study [CL]

Posted on by

http://arxiv.org/abs/2106.09499


The Maximum Entropy Spectral Analysis (MESA) method, developed by Burg, provides a powerful tool to perform spectral estimation of a time-series. The method relies on a Jaynes’ maximum entropy principle and provides the means of inferring the spectrum of a stochastic process in terms of the coefficients of some autoregressive process AR($p$) of order $p$. A closed form recursive solution provides an estimate of the autoregressive coefficients as well as of the order $p$ of the process. We provide a ready-to-use implementation of the algorithm in the form of a python package \texttt{memspectrum}. We characterize our implementation by performing a power spectral density analysis on synthetic data (with known power spectral density) and we compare different criteria for stopping the recursion. Furthermore, we compare the performance of our code with the ubiquitous Welch algorithm, using synthetic data generated from the released spectrum by the LIGO-Virgo collaboration. We find that, when compared to Welch’s method, Burg’s method provides a power spectral density (PSD) estimation with a systematically lower variance and bias. This is particularly manifest in the case of a little number of data points, making Burg’s method most suitable to work in this regime.

Read this paper on arXiv…

A. Martini, S. Schmidt and W. Pozzo
Fri, 18 Jun 21
28/62

Comments: 16 pages, 13 figure, submitted to A&A