Abstract
Principal components analysis (PCA) is the optimal linear encoder of data. Sparse linear encoders (e.g., sparse PCA) produce more interpretable features that can promote better generalization. (i) Given a level of sparsity, what is the best approximation to PCA? (ii) Are there efficient algorithms which can achieve this optimal combinatorial tradeoff? We answer both questions by providing the first polynomial-time algorithms to construct optimal sparse linear auto-encoders; additionally, we demonstrate the performance of our algorithms on real data.