連続平面上における基本的施設配置問題の競合学習法を応用した解法

• 岸本 達也

  東京大学大学院

書誌事項

タイトル別名

• Solution of Basic Facilities Locations Problem on A Continuous Plane Using Competitive Learning Method
• レンゾク ヘイメンジョウ ニ オケル キホンテキ シセツ ハイチ モンダイ ノ

抄録

Realizing the ultimate convenience for both users and service providers in Regional Facilities Location Planning is desired. This paper is to suggest a new solution of the Facilities Location Problem on Discrete Demand Distribution using the Competitive Learning Method of neural networks. By this method, no more data conversion such as mesh is needed, the original discrete data of demand distribution is input and processed. Thus, no information loss and errors occurs, and it enables more detailed planning which corresponds to the real demand distribution. Also in this paper, the superiority of the Competitive Learning Method as k-means algorithm is experimentally proved. Finally, the effectiveness of the suggested method is confirmed by the application study on a ideal facilities location problem in Meguro ward.

収録刊行物

• 都市計画論文集

  都市計画論文集 32 (0), 109-114, 1997-10-25

  公益社団法人 日本都市計画学会

キーワード

• 施設配置
• 最適化
• ニューラルネット
• 競合学習法
• 確率的降下法

詳細情報

• CRID: 1390001204214741376
• NII論文ID: 10004546772
• NII論文ID: 130006729571
• NII論文ID: 10007466543
• NII書誌ID: AN10276504
• DOI: 10.11361/journalcpij.32.109
• ISSN: 21850593
• ISSN: 09160647
• NDL書誌ID: 4452144
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I4452144
• Web Site: https://www.jstage.jst.go.jp/article/journalcpij/32/0/32_109/_pdf
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • NDL
  • Crossref
  • CiNii Articles
• 抄録ライセンスフラグ: 使用不可