On the Classification and Feature Relevance of Multiband Light Curves

Abstract With an expected torrent of data from the Vera C. Rubin Observatory Legacy Survey Space and Time (LSST), need for automated identification noisy sparse light curves will increase drastically. In this paper, we performed classification multiband astronomical Photometric LSST Astronomical Time-series Classification Challenge ( PLAsTiCC ) set via boosted neural nets, decision trees, a voted classifier 14 categories. order to deal with features, used wavelet decomposition together feature selection. We also ranking method using network. Our may be considered alternative random forests, which is known favor features more categories as relevant. investigated class importance nets one-versus-all approach reduces multiclass problem binary problem.