個体がドロップアウトする粒子群最適化法の検討(国外研修報告書)
書誌事項
- タイトル別名
-
- Analysis of Particle Swarm Optimization with Individual DropoutName(Reports of Overseas Activities)
抄録
In conventional particle swarm optimization( PSO), the diversity of an individual is lost with the state update. Such a decrease in diversity has been confirmed by some meta-heuristic algorithms. The proposed method keeps diversity by dropping out individuals from the network under swarm. In the meta-heuristics that conducts the search with many points, the method of maintaining the diversity of the swarm is expressed as regeneration of stagnant search individuals. For example, in Artificial Bee Colony optimization( ABC), a stagnant search individual is regenerated by a mechanism called a scout bee phase. In addition, it may be done implicitly due to confinement in the search space. Our proposed method keeps diversity by dropping out individuals from the network under swarm. The performance of the proposed method is confirmed using the CEC2013 benchmark function.
収録刊行物
-
- 日本工業大学研究報告 = Report of researches, Nippon Institute of Technology
-
日本工業大学研究報告 = Report of researches, Nippon Institute of Technology 49 (4), 96-99, 2020-03
- Tweet
詳細情報 詳細情報について
-
- CRID
- 1050565162988459648
-
- NII論文ID
- 120006822942
-
- ISSN
- 21895449
-
- 本文言語コード
- ja
-
- 資料種別
- departmental bulletin paper
-
- データソース種別
-
- IRDB
- CiNii Articles