Minimization of Transformed $L_1$ Penalty: Theory, Difference of Convex Function Algorithm, and Robust Application in Compressed Sensing

TL;DR

This paper introduces a novel non-convex penalty function called TL1 for compressed sensing, develops an efficient difference of convex algorithm with ADMM, and demonstrates its superior robustness and performance over existing methods.

Contribution

The paper proposes the TL1 penalty, analyzes its theoretical properties for exact and stable recovery, and develops the DCATL1 algorithm with convergence guarantees for compressed sensing applications.

Findings

DCATL1 outperforms other algorithms on Gaussian and DCT matrices.

Optimal parameter a=1 yields best performance.

DCATL1 is robust to sensing matrix conditioning.

Abstract

We study the minimization problem of a non-convex sparsity promoting penalty function, the transformed $l_1$ (TL1), and its application in compressed sensing (CS). The TL1 penalty interpolates $l_0$ and $l_1$ norms through a nonnegative parameter $a \in (0,+\infty)$, similar to $l_p$ with $p \in (0,1]$, and is known to satisfy unbiasedness, sparsity and Lipschitz continuity properties. We first consider the constrained minimization problem and discuss the exact recovery of $l_0$ norm minimal solution based on the null space property (NSP). We then prove the stable recovery of $l_0$ norm minimal solution if the sensing matrix $A$ satisfies a restricted isometry property (RIP). Next, we present difference of convex algorithms for TL1 (DCATL1) in computing TL1-regularized constrained and unconstrained problems in CS. The inner loop concerns an $l_1$ minimization problem on which we employ the Alternating Direction Method of Multipliers (ADMM). For the unconstrained problem, we prove convergence of DCATL1 to a stationary point satisfying the first order optimality condition. In numerical experiments, we identify the optimal value $a=1$, and compare DCATL1 with other CS algorithms on two classes of sensing matrices: Gaussian random matrices and over-sampled discrete cosine transform matrices (DCT). We find that for both classes of sensing matrices, the performance of DCATL1 algorithm (initiated with $l_1$ minimization) always ranks near the top (if not the top), and is the most robust choice insensitive to the conditioning of the sensing matrix $A$. DCATL1 is also competitive in comparison with DCA on other non-convex penalty functions commonly used in statistics with two hyperparameters.