torch_numopt.trust_region

Trust-region methods for optimization.

Trust-region algorithms compute a step by solving a subproblem within a region where the quadratic model is trusted. This module provides Cauchy point, dogleg, exact (with Lagrange multiplier), and Steihaug-Toint (CG) solvers.

Functions

create_trust_region_solver(method, ...[, solver])

Factory function for trust-region solvers.

Classes

CauchyPointTRSolver(curvature_estimator[, ...])	Cauchy point trust-region solver.
DoglegTRSolver(curvature_estimator[, solver])	Dogleg trust-region solver.
ExactTRSolver(curvature_estimator[, iters, tol])	Exact trust-region solver using the Lagrange multiplier method.
SteihaugTointTRSolver(curvature_estimator[, ...])	Steihaug-Toint conjugate gradient trust-region solver.
TrustRegionSolver(curvature_estimator[, solver])	Abstract base class for trust-region subproblem solvers.