成長階層型自己組織化マップを用いたパレート解集合の可視化

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2015/07/01

タイトル(英語): Visualization of Pareto Optimal Solution Sets using the Growing Hierarchical Self-organizing Maps

著者名: 鈴木　直人（千葉大学大学院工学研究科），岡本　卓（千葉大学大学院工学研究科），小圷　成一（千葉大学大学院工学研究科）

著者名(英語): Naoto Suzuki (Graduate School of Engineering, Chiba University), Takashi Okamoto (Graduate School of Engineering, Chiba University), Seiichi Koakutsu (Graduate School of Engineering, Chiba University)

キーワード: 多目的最適化，パレート解集合，可視化，自己組織化マップ，成長階層型自己組織化マップ　　Multi-objective optimization，Pareto optimal solution set，Visualization，Self-organizing maps，GHSOM

要約(英語): The visualization of the Pareto optimal solution set is one of important issues of the multi-objective optimization. The Pareto optimal solution visualization method using the self-organizing maps is one of promising visualization methods. This method has two shortcomings. One is that the map size has to be determined in advance. The other is that infeasible solutions can appear in the learnt maps. This paper proposes a new visualization technique using the growing hierarchical SOM (GHSOM), which is expected to solve foregoing shortcomings. This paper also proposes to introduce a symmetric transformation of maps into the learning algorithm in order to obtain easily viewable unified map. The effectiveness of the proposed method is confirmed through several numerical experiments.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.135 No.7 （2015） 特集：平成26 年電子・情報・システム部門大会

本誌掲載ページ: 908-919 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/135/7/135_908/_article/-char/ja/