DocumentCode :
2821058
Title :
Probabilistic model building GP with Belief propagation
Author :
Sato, Hiroyuki ; Hasegawa, Yohei ; Bollegala, Danushka ; Iba, Hitoshi
Author_Institution :
Dept. of Electr. Eng. & Inf. Syst., Univ. of Tokyo, Tokyo, Japan
fYear :
2012
fDate :
10-15 June 2012
Firstpage :
1
Lastpage :
8
Abstract :
Estimation of distribution algorithms (EDAs) which deal with tree structures as GP are called as probabilistic model building GPs (PMBGPs), and they show better search performance than GP in many problems. A problem of prototype tree-based method, a type of PMBGPs, is that samplings do not always generate the most probable solution, which is the individual with the highest probability and reflects a learned distribution most. This problem wastes a part of learning and increases the number of evaluations to get an optimum solution. In order to overcome this difficulty, this paper proposes a hybrid approach using Belief propagation (BP) in sampling process. BP is an inference algorithm on graphical models and can generate the most probable solution. By applying our approach to benchmark tests, we show that the proposed method is more effective than PLS alone.
Keywords :
belief networks; distributed algorithms; genetic algorithms; inference mechanisms; sampling methods; search problems; trees (mathematics); EDA; PLS; PMBGPs; belief propagation; benchmark tests; estimation of distribution algorithms; graphical models; inference algorithm; probabilistic logic sampling; probabilistic model building GP; prototype tree-based method; sampling process; search performance; tree structures; Bayesian methods; Belief propagation; Educational institutions; Estimation; Probabilistic logic; Prototypes; Standards;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Evolutionary Computation (CEC), 2012 IEEE Congress on
Conference_Location :
Brisbane, QLD
Print_ISBN :
978-1-4673-1510-4
Electronic_ISBN :
978-1-4673-1508-1
Type :
conf
DOI :
10.1109/CEC.2012.6256483
Filename :
6256483
Link To Document :
بازگشت