Accelerating Evolutionary Computation with Elite Obtained in Projected One-Dimensional Spaces - 九大コレクション | 九州大学附属図書館

検索結果一覧に戻る

＜会議発表論文＞
Accelerating Evolutionary Computation with Elite Obtained in Projected One-Dimensional Spaces

作成者	作成者名 Pei, Yan 裴, 岩所属機関所属機関名 Graduate School of Design, Kyushu University 九州大学大学院芸術工学府
作成者	著者識別子 K002333 0000-0002-1431-9472 作成者名 Takagi, Hideyuki 髙木, 英行タカギ, ヒデユキ所属機関所属機関名 Faculty of Design, Kyushu University 九州大学大学院芸術工学研究院
本文言語	英語
出版者	IEEE
発行日	2011-09-01
収録物名	International Conference on Genetic and Evolutionary Computing (ICGEC)
巻	2011
開始ページ	89
終了ページ	92
出版タイプ	Accepted Manuscript
アクセス権	open access
関連DOI	Proceedings, 2011 Fifth International Conference on Genetic and Evolutionary Computing, ICGEC 2011
関連DOI	https://doi.org/10.1109/ICGEC.2011.30
関連URI	https://doi.org/10.1109/ICGEC.2011.30
関連情報	Proceedings, 2011 Fifth International Conference on Genetic and Evolutionary Computing, ICGEC 2011
関連情報	https://doi.org/10.1109/ICGEC.2011.30
概要	We propose a method for accelerating evolutionary computation (EC) searches using an elite obtained in one-dimensional space and use benchmark functions to evaluate the proposed method. The method pro...jects individuals onto n one-dimensional spaces corresponding to each of the n searching parameter axes, approximates each landscape using Lagrange polynomial interpolation or power function least squares approximation, finds the best coordinate for the approximated shape, obtains an elite by combining the best n found coordinates, and uses the elite for the next generation of the EC. The advantage of this method is that the elite may be easily obtained thanks to their projection onto each onedimensional space and there is a higher possibility that the elite will be located near the global optimum. Experimental tests with differential evolution and eight benchmark functions show that the proposed method accelerates EC convergence significantly, especially in early generations.続きを見る
目次	Ⅰ.INTRODUCTION Ⅱ.OBTAINING ELITE FROM A REGRESSION SEARCH SPACE Ⅲ.EXPERIMENTAL EVALUATIONS Ⅳ.DISCUSSIONS Ⅴ.CONCLUSION

本文ファイル

ファイル	ファイルタイプ	サイズ	閲覧回数	説明
ICGEC2011Pei	pdf	372 KB	345

詳細

レコードID	1808904
査読有無	査読無
主題	evolutionary computation
	fitness landscape
	convergence acceleration
	dimensionality reduction
	elite combination search
ISBN	978-1-4577-0817-6
DOI	10.1109/ICGEC.2011.30
注記	2011 Fifth International Conference on Genetic and Evolutionary Computing ICGEC 2011 : 29 August - 1 September 2011 : Kinmen, Taiwan/Xiamen, China
登録日	2017.06.14
更新日	2021.10.06