Methods for High-Dimensional Nonlinear Optimization
Abstract
Stochastic nonlinear optimization problems arise in a wide variety of data science applications. We propose a general algorithmic framework for solving optimization problems that are highdimensional; nonlinear (and possibly ill conditioned); nonconvex (and possibly ill posed). The proposed methods employ new sampling techniques, can take advantage of parallel architectures, and are supported by a convergence theory.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 23, 2018
- Source ID
- N000141812098
Entities
People
- Jorge Nocedal
Organizations
- Northwestern University
- Office of Naval Research
- United States Navy