Hybridizing Cartesian Genetic Programming and Harmony Search for adaptive feature construction in supervised learning problems

详细信息    查看全文

• 作者：Andoni Elola^a ; Javier Del Ser^a ; ^b ; ^c ; ^{javier.delser@tecnalia.com} ; Miren Nekane Bilbao^a ; Cristina Perfecto^a ; Enrique Alexandre^d ; Sancho Salcedo-Sanz^d
• 关键词：Feature construction ; Supervised learning ; Cartesian Genetic Programming ; Harmony Search
• 刊名：Applied Soft Computing
• 出版年：2017
• 出版时间：March 2017
• 年：2017
• 卷：52
• 期：Complete
• 页码：760-770
• 全文大小：2001 K
• 卷排序：52

文摘

We present a new iterative feature construction approach for supervised learning model based on the meta-heuristic Harmony Search (HS) algorithm and Cartesian Genetic Programming. We propose a novel method to incorporate soft information about the relevance of the constructed features in the HS algorithm so as to enhance its convergence. The performance of the proposed scheme is assessed over datasets from the literature, with promising results that support its suitability to deal with legacy datasets.