EFFICIENT NATURAL GRADIENT DESCENT METHODS FOR LARGE-SCALE PDE-BASED OPTIMIZATION PROBLEMS

doi:10.1137/22M1477805

EFFICIENT NATURAL GRADIENT DESCENT METHODS FOR LARGE-SCALE PDE-BASED OPTIMIZATION PROBLEMS

2023-07-10

6

OA

AI

DOI：10.1137/22M1477805

Original

Original request for help

Save

Abstract

En 中文

We propose efficient numerical schemes for implementing the natural gradient descent (NGD) for a broad range of metric spaces with applications to PDE-based optimization problems. Our technique represents the natural gradient direction as a solution to a standard least-squares problem. Hence, instead of calculating, storing, or inverting the information matrix directly, we apply efficient methods from numerical linear algebra. We treat both scenarios where the Jacobian, i.e., the derivative of the state variable with respect to the parameter, is either explicitly known or implicitly given through constraints. We can thus reliably compute several natural NGDs for a large-scale parameter space. In particular, we are able to compute Wasserstein NGD in thousands of dimensions, which was believed to be out of reach. Finally, our numerical results shed light on the qualitative differences between the standard gradient descent and various NGD methods based on different metric spaces in nonconvex optimization problems.

Keywords:

natural gradient

constrained optimization

least-squares method

gradient flow

inverse problem

EFFICIENT NATURAL GRADIENT DESCENT METHODS FOR LARGE-SCALE PDE-BASED OPTIMIZATION PROBLEMS

Abstract

Journal

Organization