In the last decade, over a million stars were monitored to detect transiting planets. Manual interpretation of potential exoplanet candidates is labor intensive and subject to human error, the results of which are difficult to quantify. Here we present a new method of detecting exoplanet candidates in large planetary search projects which, unlike current methods uses a neural network. Neural networks, also called “deep learning” or “deep nets”, are a state of the art machine learning technique designed to give a computer perception into a specific problem by training it to recognize patterns. Unlike past transit detection algorithms deep nets learn to recognize planet features instead of relying on hand-coded metrics that humans perceive as the most representative. Our deep learning algorithms are capable of detecting Earth-like exoplanets in noisy time-series data with 99$\%$ accuracy compared to a 73$\%$ accuracy using least-squares. For planet signals smaller than the noise we devise a method for finding periodic transits using a phase folding technique that yields a constraint when fitting for the orbital period. Deep nets are highly generalizable allowing data to be evaluated from different time series after interpolation. We validate our deep net on light curves from the Kepler mission and detect periodic transits similar to the true period without any model fitting.
K. Pearson, L. Palafox and C. Griffith
Thu, 15 Jun 17
Comments: Article has undergone one revision at MNRAS. Additional comments welcome