...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Evolutionary computation >Selection Heuristics on Semantic Genetic Programming for Classification Problems
【24h】

Selection Heuristics on Semantic Genetic Programming for Classification Problems

机译：Selection Heuristics on Semantic Genetic Programming for Classification Problems

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相关主题

摘要

Individual semantics have been used for guiding the learning process of Genetic Programming.Novel genetic operators and different ways of performing parent selectionhave been proposed with the use of semantics. The latter is the focus of this contributionby proposing three heuristics for parent selection that measure the similarityamong individuals’ semantics for choosing parents that enhance the addition, NaiveBayes, and Nearest Centroid. To the best of our knowledge, this is the first time thatfunctions’ properties are used for guiding the learning process. As the heuristics werecreated based on the properties of these functions, we apply them only when they areused to create offspring. The similarity functions considered are the cosine similarity,Pearson’s correlation, and agreement.We analyze these heuristics’ performance againstrandom selection, state-of-the-art selection schemes, and 18 classifiers, including automachine-learning techniques, on 30 classification problems with a variable number ofsamples, variables, and classes. The result indicated that the combination of parent selectionbased on agreement and random selection to replace an individual in the populationproduces statistically better results than the classical selection and state-of-theartschemes, and it is competitive with state-of-the-art classifiers. Finally, the code isreleased as open-source software.

著录项

获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号