ℓp-Regularized Least Squares (0

Abstract

This paper elucidates the underlying structures of ℓ_p-regularized least squares problems in the nonconvex case of 0 <; p <; 1. The difference between two formulations is highlighted (which does not occur in the convex case of p = 1): 1) an ℓ_p-constrained optimization (P_c^p) and 2) an ℓ_p-penalized (unconstrained) optimization (L_λ^p). It is shown that the solution path of (L_λ^p) is discontinuous and also a part of the solution path of (P_c^p). As an alternative to the solution path, a critical path is considered, which is a maximal continuous curve consisting of critical points. Critical paths are piecewise smooth, as can be seen from the viewpoint of the variational method, and generally contain non-optimal points, such as saddle points and local maxima as well as global/local minima. Our study reveals multiplicity (non-monotonicity) in the correspondence between the regularization parameters of (P_c^p) and (L_λ^p ). Two particular paths of critical points connecting the origin and an ordinary least squares (OLS) solution are studied further. One is a main path starting at an OLS solution, and the other is a greedy path starting at the origin. Part of the greedy path can be constructed with a generalized Minkowskian gradient. This paper of greedy path leads to a nontrivial close-link between the optimization problem of ℓ_p-regularized least squares and the greedy method of orthogonal matching pursuit.