TOPICS IN MODERN NONCONVEX NONDIFFERENTIABLE OPTIMIZATION

Abstract

The research is to develop new theory and algorithms for solving important classes of optimization problems that are nonconvex, nondifferentiable, and-or nonregular. These problems, collectively referred to as non -problems, are considered the most difficult of mathematical programs since they do not enjoy desirable properties that allow for their efficient solving. Locally optimal solutions are typically not globally optimal, and the lack of differentiability precludes the usage of efficient gradient-based methods. The effort builds on past successes of the PI, and the anticipated results are expected to significantly advance core optimization research and to broaden its impact into many areas such as statistical learning, artificial intelligence, bilevel decision making, and robustification. Theoretical and algorithmic advances for the problems under investigation can greatly enhance the DoD’s ability to efficiently solve a broad array of decision-making problems, including those for which the input parameters are known with certainty as well as for those with stochastic variations.

Document Details

Document Type
DoD Grant Award
Publication Date
Mar 07, 2023
Source ID
FA95502210045

Entities

People

  • Jong-shi Pang

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Southern California

Tags

Readers

  • Operations Research
  • Research Science/Academic Research
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • AI & ML - Machine Learning Algorithms