An empirical study of functional complexity as an indicator of overfitting in Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.7630

@InProceedings{trujillo:2011:EuroGP,

• author = "Leonardo Trujillo and Sara Silva and Pierrick Legrand and Leonardo Vanneschi",
• title = "An empirical study of functional complexity as an indicator of overfitting in Genetic Programming",
• booktitle = "Proceedings of the 14th European Conference on Genetic Programming, EuroGP 2011",
• year = "2011",
• month = "27-29 " # apr,
• editor = "Sara Silva and James A. Foster and Miguel Nicolau and Mario Giacobini and Penousal Machado",
• series = "LNCS",
• volume = "6621",
• publisher = "Springer Verlag",
• address = "Turin, Italy",
• pages = "262--273",
• organisation = "EvoStar",
• keywords = "genetic algorithms, genetic programming: poster",
• isbn13 = "978-3-642-20406-7",
• DOI = "doi:10.1007/978-3-642-20407-4_23",
• abstract = "Recently, it has been stated that the complexity of a solution is a good indicator of the amount of overfitting it incurs. However, measuring the complexity of a program, in Genetic Programming, is not a trivial task. In this paper, we study the functional complexity and how it relates with overfitting on symbolic regression problems. We consider two measures of complexity, Slope-based Functional Complexity, inspired by the concept of curvature, and Regularity-based Functional Complexity based on the concept of Holderian regularity. In general, both complexity measures appear to be poor indicators of program overfitting. However, results suggest that Regularity-based Functional Complexity could provide a good indication of overfitting in extreme cases.",
• notes = "Part of \cite{Silva:2011:GP} EuroGP'2011 held in conjunction with EvoCOP2011 EvoBIO2011 and EvoApplications2011",

}

Genetic Programming entries for Leonardo Trujillo Sara Silva Pierrick Legrand Leonardo Vanneschi

Citations