Algorithmic Strategies for Non-convex Optimization in Sparse Learning

We consider optimization formulations with non-convex regularization that are natural for learning sparse linear models. There are two approaches to this problem: 1. Heuristic methods such as gradient