Indirectly Parameterized Concrete Autoencoders

arXiv:2403.00563v1 Announce Type: new
Abstract: Feature selection is a crucial task in settings where data is high-dimensional or acquiring the full set of features is costly. Recent developments in neural network-based embedded feature selection show promising results across a wide range of applications. Concrete Autoencoders (CAEs), considered state-of-the-art in embedded feature selection, may struggle to achieve stable joint optimization, hurting their training time and generalization. In this work, we identify that this instability is correlated with the CAE learning duplicate …

abstract applications art arxiv autoencoders concrete cs.lg data embedded feature features feature selection network neural network results set show state stat.ml struggle type