Reinforcement learning for adaptive operator selection in memetic search applied to Quadratic Assignment Problem

Reinforcement learning for adaptive operator selection in memetic search applied to Quadratic Assignment Problem

Memetic search is well known as one of the state-of-the-art metaheuristics for finding high-quality solutions to NP-hard problems. Its performance is often attributable to appropriate design, including the choice of its operators. In this paper, we propose a Markov Decision Process model for the sel...

Full description

Saved in:
Bibliographic Details
Main Authors: HANDOKO, Stephanus Daniel, Nguyen, Duc Thien, YUAN, Zhi, LAU, Hoong Chuin
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2014
Subjects:
Artificial Intelligence and Robotics
Operations Research, Systems Engineering and Industrial Engineering
Online Access:https://ink.library.smu.edu.sg/sis_research/2666
https://ink.library.smu.edu.sg/context/sis_research/article/3666/viewcontent/ReinforcementLearningAdaptiveOperSelQAP_2014_GECCO.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
Description
Summary:Memetic search is well known as one of the state-of-the-art metaheuristics for finding high-quality solutions to NP-hard problems. Its performance is often attributable to appropriate design, including the choice of its operators. In this paper, we propose a Markov Decision Process model for the selection of crossover operators in the course of the evolutionary search. We solve the proposed model by a Q-learning method. We experimentally verify the efficacy of our proposed approach on the benchmark instances of Quadratic Assignment Problem.

Similar Items