Efficient nonconvex sparse group feature selection via continuous and discrete optimization

Sparse feature selection has proven to be effective in analyzing high-dimensional data. While promising, most existing works apply convex methods, which may be suboptimal in terms of the accuracy o...