An Adaptive Sampling Augmented Lagrangian Method for Stochastic Optimization with Deterministic Constraints
Preprint
Abstract
The primary goal of this paper is to provide an efficient solution algorithm based on the augmented Lagrangian framework for optimization problems with a stochastic objective function and deterministic constraints. Our main contribution is combining the augmented Lagrangian framework with adaptive sampling, resulting in an efficient optimization methodology validated with practical examples. To achieve the presented efficiency, we consider inexact solutions for the augmented Lagrangian subproblems, and through an adaptive sampling mechanism, we control the variance in the gradient estimates. Furthermore, we analyze the theoretical performance of the proposed scheme by showing equivalence to a gradient descent algorithm on a Moreau envelope function, and we prove sublinear convergence for convex objectives and linear convergence for strongly convex objectives with affine equality constraints. The worst-case sample complexity of the resulting algorithm, for an arbitrary choice of penalty parameter in the augmented Lagrangian function, is $\mathcal{O}(\epsilon^{-3-\delta})$, where $\epsilon > 0$ is the expected error of the solution and $\delta > 0$ is a user-defined parameter. If the penalty parameter is chosen to be $\mathcal{O}(\epsilon^{-1})$, we demonstrate that the result can be improved to $\mathcal{O}(\epsilon^{-2})$, which is competitive with the other methods employed in the literature. Moreover, if the objective function is strongly convex with affine equality constraints, we obtain $\mathcal{O}(\epsilon^{-1}\log(1/\epsilon))$ complexity. Finally, we empirically verify the performance of our adaptive sampling augmented Lagrangian framework in machine learning optimization and engineering design problems, including topology optimization of a heat sink with environmental uncertainty.
Keywords
All Related Versions
This publication has 47 references indexed in Scilit:
- Inequality constrained stochastic nonlinear optimization via active-set sequential quadratic programmingMathematical Programming, 2023
- An adaptive stochastic sequential quadratic programming with differentiable exact augmented lagrangiansMathematical Programming, 2022
- A primal–dual algorithm for risk minimizationMathematical Programming, 2021
- Global convergence rate analysis of unconstrained optimization methods based on probabilistic modelsMathematical Programming, 2017
- Length scale and manufacturability in density-based topology optimizationArchive of Applied Mechanics, 2016
- Iteration-complexity of first-order augmented Lagrangian methods for convex programmingMathematical Programming, 2015
- An adaptive augmented Lagrangian method for large-scale constrained optimizationMathematical Programming, 2014
- On the similarities between micro/nano lithography and topology optimization projection methodsStructural and Multidisciplinary Optimization, 2013
- Filters in topology optimization based on Helmholtz‐type differential equationsInternational Journal for Numerical Methods in Engineering, 2010
- Level set based robust shape and topology optimization under random field uncertaintiesStructural and Multidisciplinary Optimization, 2009