Genetic Programming for large tree optimization with considering inheritance and acquisition of traits

2017 Fiscal Year Final Research Report

• Project/Area Number: 26330290
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Soft computing
• Research Institution: Kansai University
• Principal Investigator: Hanada Yoshiko 関西大学, システム理工学部, 准教授 (30511711)
• Co-Investigator(Kenkyū-buntansha): 小野 景子 龍谷大学, 理工学部, 講師 (80550235)
• Project Period (FY): 2014-04-01 – 2018-03-31
• Keywords: 遺伝的プログラミング / 木構造 / 局所探索 / 初期化

Outline of Final Research Achievements

The purpose of this study is to develop a new genetic programming for large tree optimization in order to solve the design problem consisting of variables expressed in tree structure such as expressions and program generation for autonomous control. The achieved works of this study are: (1) Improvement of a multi step crossover that has high trait integrity, (2) Development of an initialization scheme based on preferable trait estimation by past search information, and (3) Development of a search scheme for obtaining new traits. We showed the effectiveness and application guidelines of the proposed method mainly through the performance analysis using the function regression problem.

Free Research Field

最適化