Coordinate Linear Variance Reduction for Generalized Linear Programming
arXiv:2111.01842
Abstract
We study a class of generalized linear programs (GLP) in a large-scale setting, which includes simple, possibly nonsmooth convex regularizer and simple convex set constraints. By reformulating (GLP) as an equivalent convex-concave min-max problem, we show that the linear structure in the problem can be used to design an efficient, scalable first-order algorithm, to which we give the name \emph{Coordinate Linear Variance Reduction} (\textsc{clvr}; pronounced "clever"). \textsc{clvr} yields improved complexity results for (GLP) that depend on the max row norm of the linear constraint matrix in (GLP) rather than the spectral norm. When the regularization terms and constraints are separable, \textsc{clvr} admits an efficient lazy update strategy that makes its complexity bounds scale with the number of nonzero elements of the linear constraint matrix in (GLP) rather than the matrix dimensions. On the other hand, for the special case of linear programs, by exploiting sharpness, we propose a restart scheme for \textsc{clvr} to obtain empirical linear convergence. Then we show that Distributionally Robust Optimization (DRO) problems with ambiguity sets based on both -divergence and Wasserstein metrics can be reformulated as (GLPs) by introducing sparsely connected auxiliary variables. We complement our theoretical guarantees with numerical experiments that verify our algorithm's practical effectiveness, in terms of wall-clock time and number of data passes.
39 pages, NeurIPS 2022
References in corpus (6)
- Faster First-Order Primal-Dual Methods for Linear Programming using Restarts and Sharpness
- Stochastic Variance Reduction for Variational Inequality Methods
- Practical Large-Scale Linear Programming using Primal-Dual Hybrid Gradient
- A First-Order Algorithmic Framework for Wasserstein Distributionally Robust Logistic Regression
- Fast Distributionally Robust Learning with Variance Reduced Min-Max Optimization
- Random extrapolation for primal-dual coordinate descent