On proximal gradient mapping and its minimization in norm via potential function-based acceleration
Optimization and Control
2022-12-15 v1
Abstract
The proximal gradient descent method, well-known for composite optimization, can be completely described by the concept of proximal gradient mapping. In this paper, we highlight our previous two discoveries of proximal gradient mapping--norm monotonicity and refined descent, with which we are able to extend the recently proposed potential function-based framework from gradient descent to proximal gradient descent.
Cite
@article{arxiv.2212.07149,
title = {On proximal gradient mapping and its minimization in norm via potential function-based acceleration},
author = {Beier Chen and Hui Zhang},
journal= {arXiv preprint arXiv:2212.07149},
year = {2022}
}
Comments
16 pages