Current browse context:
cs.LG
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: A Unified Framework for Constructing Nonconvex Regularizations
(Submitted on 11 Jun 2021)
Abstract: Over the past decades, many individual nonconvex methods have been proposed to achieve better sparse recovery performance in various scenarios. However, how to construct a valid nonconvex regularization function remains open in practice. In this paper, we fill in this gap by presenting a unified framework for constructing the nonconvex regularization based on the probability density function. Meanwhile, a new nonconvex sparse recovery method constructed via the Weibull distribution is studied.
Link back to: arXiv, form interface, contact.