## Sparse/Robust Estimation and Kalman Smoothing with Nonsmooth Log-Concave Densities: Modeling, Computation, and Theory

Piecewise linear quadratic (PLQ) penalties play a crucial role in many applications, including machine learning, robust statistical inference, sparsity promotion, and inverse problems such as Kalman smoothing. Well known examples of PLQ penalties include the l2, Huber, l1 and Vapnik losses. This paper builds on a dual representation for PLQ penalties known from convex analysis. … Read more