High-Order Moreau Envelope in the Nonconvex Setting: Framework and Algorithms
Offered By: Erwin Schrödinger International Institute for Mathematics and Physics (ESI) via YouTube
Course Description
Overview
Explore a 31-minute conference talk on high-order Moreau envelope in nonconvex optimization, presented by Masoud Ahookhosh at the Erwin Schrödinger International Institute for Mathematics and Physics. Delve into the introduction of high-order proximal operator (HOPE) and high-order Moreau envelope (HOME) in nonconvex settings. Examine the fundamental properties of HOPE and HOME, including single-valuedness and differential properties under conditions like prox-regularity and weak-convexity. Investigate the Hölder continuity of HOPE and gradient of HOME. Learn about the development of inexact proximal point and inexact gradient methods for nonconvex optimization, including their convergence analysis and linear convergence under Kurdyka-Łojasiewicz conditions. Gain insights from preliminary numerical results that validate the theoretical foundations presented in this talk, which was part of the "One World Optimization Seminar in Vienna" workshop held at ESI in June 2024.
Syllabus
Masoud Ahookhosh - High-order Moreau envelope in the nonconvex setting: Framework and algorithms
Taught by
Erwin Schrödinger International Institute for Mathematics and Physics (ESI)
Related Courses
Deep Learning for Natural Language ProcessingUniversity of Oxford via Independent Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
DeepLearning.AI via Coursera Deep Learning Part 1 (IITM)
Indian Institute of Technology Madras via Swayam Deep Learning - Part 1
Indian Institute of Technology, Ropar via Swayam Logistic Regression with Python and Numpy
Coursera Project Network via Coursera