Proposal of an Interactive Evolutionary Algorithm with No Explicit Numerical Evaluation of Solutions by a Human User

无需人类用户对解决方案进行显式数值评估的交互式进化算法的提案

• 批准号: 23650119
• 负责人: ISHIBUCHI Hisao
• 金额: $ 2.25万
• 依托单位: Osaka Prefecture University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Challenging Exploratory Research
• 财政年份: 2011
• 资助国家: 日本
• 起止时间: 2011 至 2013
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-23650119/
• 关键词: 進化計算; 対話型進化計算; 多目的最適化; 利用者; 選好情報

We proposed an interactive evolutionary computation algorithm with the minimum workload of a human user in solution evaluation. In the proposed algorithm, a single solution is presented to the user at a time. The user is supposed to evaluate the presented solution in order to decide whether the presented solution is better than the previous one or not. After the evaluation of each solution, the proposed algorithm updates a set of candidate solutions and generates a new solution to be presented to the user. The proposed algorithm is able to show a single candidate solution after a pre-specified number of iterations of solution evaluation by the user. That is, the set of candidate solutions can be decreased to a single solution after an arbitrarily specified number of iterations of solution evaluation.

我们提出了一种交互式进化计算算法的最小工作量的人类用户在解决方案的评价。在所提出的算法中，一次向用户呈现单个解决方案。用户应该评估所呈现的解决方案，以决定所呈现的解决方案是否优于先前的解决方案。每个解决方案的评估后，该算法更新一组候选解决方案，并生成一个新的解决方案呈现给用户。该算法是能够显示一个单一的候选解决方案后，由用户的解决方案评估的迭代次数预先指定。也就是说，在任意指定次数的解评估迭代之后，候选解的集合可以减少到单个解。

项目成果

Problem formulation of interactive evolutionary computation with minimum requirement for human user's fitness evaluation ability

对人类用户适应度评估能力最低要求的交互式进化计算问题表述

• 发表时间: 2012
• 作者: H. Ishibuchi;K. Hoshino;and Y. Nojima
• 通讯作者: and Y. Nojima

Archive management in interactive evolutionary computation with minimum requirement for human user's fitness evaluation ability

对人类用户适应度评估能力要求最低的交互式进化计算中的档案管理

• DOI: 10.1007/978-3-319-07173-2_31
• 发表时间: 2014
• 期刊: Lecture Notes in Computer Science
• 作者: H. Ishibuchi;T. Sudo;and Y. Nojima
• 通讯作者: and Y. Nojima

Difficulty in Evolutionary Multiobjective Optimization of Discrete Objective Functions with Different Granularities

• DOI: 10.1007/978-3-642-37140-0_20
• 发表时间: 2013-03
• 作者: H. Ishibuchi;M. Yamane;Y. Nojima

Effects of discrete objective functions with different granularities on the search behavior of EMO algorithms

• DOI: 10.1145/2330163.2330232
• 发表时间: 2012-07
• 作者: H. Ishibuchi;M. Yamane;Y. Nojima

Interactive (1+1) evolutionary strategy with one-fifth success rule

五分之一成功规则的交互式（1 1）进化策略

• 发表时间: 2014
• 作者: T. Sudo;K. Ueba;Y. Nojima;and H. Ishibuchi
• 通讯作者: and H. Ishibuchi