Annealed Sparsity via Adaptive and Dynamic Shrinking

Sparse learning has received tremendous amount of interest in high-dimensional data analysis due to its model interpretability and the low-computational cost. Among the various techniques, adaptive l1