http://arxiv.org/abs/2211.07489
Aims. We present the application of a fully connected neural network for galaxy merger identification using exclusively photometric information. Our purpose is not only to test the method’s efficiency but also to understand what merger properties the neural network can learn and what their physical interpretation is. Methods. We created a class-balanced training dataset of 5860 galaxies split into mergers and non-mergers. The galaxy observations come from SDSS DR6 and were visually identified in Galaxy Zoo. The 2$\,$930 mergers were selected from known SDSS mergers and the respective non-mergers are the closest match in both redshift and $r$ magnitude. The NN architecture was built by testing a different number of layers with different sizes and variations of the dropout rate. We compare input spaces constructed using the five SDSS filters $u$, $g$, $r$, $i$ and $z$; combinations of bands, colours and their errors; six magnitude types and variations of input normalization. Results. It was found that the fibre magnitude errors contribute the most to the training accuracy. Studying the parameters from which they are calculated, we showed that the input space built from the sky error background in the five SDSS bands alone leads to 92.64 $\pm$ 0.15 \% training accuracy. We also found that the input normalization, i.e., how the data are presented to the NN, has a significant effect on the training performance. Conclusions. We conclude that, from all the SDSS photometric information, the sky error background is the most sensitive to merging processes. This finding is supported by an analysis of its 5-band feature space by means of data visualization. Moreover, studying the plane of the $g$ and $r$ sky error bands shows that a decision boundary line is enough to achieve a 91.59\% accuracy.
L. Suelves, W. Pearson and A. Pollo
Tue, 15 Nov 22
31/103
Comments: Accepted in Astronomy and Astrophysics (A&A) on 05/11/2022