Sequential Monte Carlo-based multi-objective optimization

Juan Pablo Sáenz, Nurcin Celik

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Multi-objective optimization problems are often found in everyday life, such as in the tradeoff between the cost and quality of a product. As the simultaneous multiple objectives may conflict with each other, the optimization problem is very challenging, and there rarely is a single global optimum. There are two common approaches for multi-objective optimization problems. The first focuses on transforming the problem into a single objective problem through the use of normalization or combination techniques, whereas the second uses multi-objective evolutionary algorithms that construct a non-dominated solution set. As an alternative, in this paper we present a sequential Monte Carlo algorithm with two-stage sampling for multi-objective optimization problems. In the first stage, sampling is executed from within the non-dominated solution set generated by the algorithm; while in the second stage, sampling is performed from within the extreme points of the non-dominated solution set and some of the closest extreme points of the search space. The proposed approach has been benchmarked against well-known multi-objective optimization algorithms. It has performed better than the alternatives in problems where the Pareto front presents convexity, or multimodality; while producing promising results in instances with discontinuity along the Pareto front.

Original languageEnglish
Title of host publicationIIE Annual Conference and Expo 2013
PublisherInstitute of Industrial Engineers
Pages11-20
Number of pages10
StatePublished - Jan 1 2013
EventIIE Annual Conference and Expo 2013 - San Juan, Puerto Rico
Duration: May 18 2013May 22 2013

Other

OtherIIE Annual Conference and Expo 2013
CountryPuerto Rico
CitySan Juan
Period5/18/135/22/13

Keywords

  • Multi-objective optimization
  • Pareto optimality
  • Particle filtering
  • Sequential Monte Carlo

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering

Fingerprint Dive into the research topics of 'Sequential Monte Carlo-based multi-objective optimization'. Together they form a unique fingerprint.

Cite this