Efficiently Solving Stochastic Mixed-Integer Problems combining Gauss–Siedel and Penalty-Based methods

Abstract

In this talk, we will present the development of a novel decomposition approach for mixed-integer stochastic programming (SMIP) problems that is inspired by the combination of penalty-based Lagrangian and block Gauss–Seidel methods (the PBGS method). In this sense, we will present the technical aspects associated with the development of PBGS, which in turn focus on exploiting in a computationally efficient manner the inherent decomposable structure that SMIPs present. The performance of the proposed method is compared with the classical Progressive Hedging method (PH), which also can be viewed as a Lagrangian relaxation-based method for obtaining solutions for SMIP. We will also present extensive numerical results performed using several instances from the literature that illustrate the efficiency of the proposed method in terms of computational performance and solution quality.

Date
Feb 23, 2019 2:45 PM
Location
Adelaide, Australia
Fabricio Oliveira
Fabricio Oliveira
Associate Professor of Operational Research

Fabricio Oliveira is an Associate Professor of Operational Research in the Department of Mathematics and Systems Analysis.