An extended probabilistic model building genetic network programming using both of good and bad individuals

Xianneng Li, Shingo Mabu, Kotaro Hirasawa

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Classical estimation of distribution algorithms (EDAs) generally use truncation selection to estimate the distribution of the good individuals while ignoring the bad ones. However, various researches in evolutionary algorithms (EAs) have reported that the bad individuals may affect and help solving the problem. This paper proposes a new method to use the bad individuals by studying the substructures rather than the entire individual structures to solve reinforcement learning (RL) problems, which generally factorize their entire solutions to the sequences of state-action pairs. This work was studied in a recent graph-based EDA named probabilistic model building genetic network programming (PMBGNP), which could solve RL problems successfully, to propose an extended PMBGNP. The effectiveness of this work is verified in an RL problem, namely robot control. Compared to other related work, results show that the proposed method can significantly speed up the evolution efficiency.

Original languageEnglish
Pages (from-to)339-347
Number of pages9
JournalIEEJ Transactions on Electrical and Electronic Engineering
Volume8
Issue number4
DOIs
Publication statusPublished - 2013 Jul

Fingerprint

Reinforcement learning
Evolutionary algorithms
Robots
Statistical Models

Keywords

  • Bad individuals
  • Estimation of distribution algorithms (EDAs)
  • Probabilistic model building genetic network programming
  • Probabilistic modeling
  • Reinforcement learning

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

An extended probabilistic model building genetic network programming using both of good and bad individuals. / Li, Xianneng; Mabu, Shingo; Hirasawa, Kotaro.

In: IEEJ Transactions on Electrical and Electronic Engineering, Vol. 8, No. 4, 07.2013, p. 339-347.

Research output: Contribution to journalArticle

@article{9e0606840e7345958f7c47fe03a87ff4,
title = "An extended probabilistic model building genetic network programming using both of good and bad individuals",
abstract = "Classical estimation of distribution algorithms (EDAs) generally use truncation selection to estimate the distribution of the good individuals while ignoring the bad ones. However, various researches in evolutionary algorithms (EAs) have reported that the bad individuals may affect and help solving the problem. This paper proposes a new method to use the bad individuals by studying the substructures rather than the entire individual structures to solve reinforcement learning (RL) problems, which generally factorize their entire solutions to the sequences of state-action pairs. This work was studied in a recent graph-based EDA named probabilistic model building genetic network programming (PMBGNP), which could solve RL problems successfully, to propose an extended PMBGNP. The effectiveness of this work is verified in an RL problem, namely robot control. Compared to other related work, results show that the proposed method can significantly speed up the evolution efficiency.",
keywords = "Bad individuals, Estimation of distribution algorithms (EDAs), Probabilistic model building genetic network programming, Probabilistic modeling, Reinforcement learning",
author = "Xianneng Li and Shingo Mabu and Kotaro Hirasawa",
year = "2013",
month = "7",
doi = "10.1002/tee.21864",
language = "English",
volume = "8",
pages = "339--347",
journal = "IEEJ Transactions on Electrical and Electronic Engineering",
issn = "1931-4973",
publisher = "John Wiley and Sons Inc.",
number = "4",

}

TY - JOUR

T1 - An extended probabilistic model building genetic network programming using both of good and bad individuals

AU - Li, Xianneng

AU - Mabu, Shingo

AU - Hirasawa, Kotaro

PY - 2013/7

Y1 - 2013/7

N2 - Classical estimation of distribution algorithms (EDAs) generally use truncation selection to estimate the distribution of the good individuals while ignoring the bad ones. However, various researches in evolutionary algorithms (EAs) have reported that the bad individuals may affect and help solving the problem. This paper proposes a new method to use the bad individuals by studying the substructures rather than the entire individual structures to solve reinforcement learning (RL) problems, which generally factorize their entire solutions to the sequences of state-action pairs. This work was studied in a recent graph-based EDA named probabilistic model building genetic network programming (PMBGNP), which could solve RL problems successfully, to propose an extended PMBGNP. The effectiveness of this work is verified in an RL problem, namely robot control. Compared to other related work, results show that the proposed method can significantly speed up the evolution efficiency.

AB - Classical estimation of distribution algorithms (EDAs) generally use truncation selection to estimate the distribution of the good individuals while ignoring the bad ones. However, various researches in evolutionary algorithms (EAs) have reported that the bad individuals may affect and help solving the problem. This paper proposes a new method to use the bad individuals by studying the substructures rather than the entire individual structures to solve reinforcement learning (RL) problems, which generally factorize their entire solutions to the sequences of state-action pairs. This work was studied in a recent graph-based EDA named probabilistic model building genetic network programming (PMBGNP), which could solve RL problems successfully, to propose an extended PMBGNP. The effectiveness of this work is verified in an RL problem, namely robot control. Compared to other related work, results show that the proposed method can significantly speed up the evolution efficiency.

KW - Bad individuals

KW - Estimation of distribution algorithms (EDAs)

KW - Probabilistic model building genetic network programming

KW - Probabilistic modeling

KW - Reinforcement learning

UR - http://www.scopus.com/inward/record.url?scp=84879275907&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84879275907&partnerID=8YFLogxK

U2 - 10.1002/tee.21864

DO - 10.1002/tee.21864

M3 - Article

AN - SCOPUS:84879275907

VL - 8

SP - 339

EP - 347

JO - IEEJ Transactions on Electrical and Electronic Engineering

JF - IEEJ Transactions on Electrical and Electronic Engineering

SN - 1931-4973

IS - 4

ER -