Id: arxiv/2603.09722 Url: https://arxiv.org/abs/2603.09722 Tags: math.FA, cs.IT, math.CA Score: 0 Title: A Novel Two-Parameter Penalty: Relaxation Degree Analysis and Sparse Signal Recovery Domain: math.FA Source: arxiv Authors: Ziwei Li, Wengu Chen, Huanmin Ge, Dachun Yang Reviews: 0 Subject: math.FA Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-03-10 Page: https://delphi.science/r/arxiv/2603.09722 In this article, we introduce a nonconvex two-parameter penalty function $P_{a,p}$, parameterized by $a\in(0,\infty)$ and $p\in(0,1]$, and the relaxation degree RD$_P$ for a separable nonconvex penalty function $P$. Based on $P_{a,p}$, we further propose the $P_{a,p}$ minimization framework for sparse signal recovery. This framework generalizes the TL1 minimization model established by S. Zhang and J. Xin (corresponding to the special case $p=1$) and provides a unified and flexible family of nonconvex penalty functions for sparse signal recovery. Using the sparse convex-combination…