Risk Bounds for Learning Multiple Components with Permutation-Invariant Losses

Fabien Lauer

This paper proposes a simple approach to derive efficient error bounds for learning multiple components with sparsity-inducing regularization. We show that for such regularization…