A CNN adapted to time series for the classification of Supernovae
This work addresses the challenge of analyzing large astronomical datasets for cosmologists, offering improved classification methods for supernovae, though it appears incremental as it builds on existing CNN approaches.
The authors tackled the problem of classifying supernovae, specifically distinguishing Type Ia from non-Ia supernovae, by presenting two Convolutional Neural Networks (CNNs) that defeat the current state-of-the-art, with one adapted to time series for light-curves and another using a Siamese CNN for sparse and limited data.
Cosmologists are facing the problem of the analysis of a huge quantity of data when observing the sky. The methods used in cosmology are, for the most of them, relying on astrophysical models, and thus, for the classification, they usually use a machine learning approach in two-steps, which consists in, first, extracting features, and second, using a classifier. In this paper, we are specifically studying the supernovae phenomenon and especially the binary classification "I.a supernovae versus not-I.a supernovae". We present two Convolutional Neural Networks (CNNs) defeating the current state-of-the-art. The first one is adapted to time series and thus to the treatment of supernovae light-curves. The second one is based on a Siamese CNN and is suited to the nature of data, i.e. their sparsity and their weak quantity (small learning database).