论文信息 - Natural Selection at Work: An Accelerated Evolutionary Computing Approach to Predictive Model Selection

Natural Selection at Work: An Accelerated Evolutionary Computing Approach to Predictive Model Selection

We implement genetic algorithm based predictive model building as an alternative to the traditional stepwise regression. We then employ the Information Complexity Measure (ICOMP) as a measure of model fitness instead of the commonly used measure of R-square. Furthermore, we propose some modifications to the genetic algorithm to increase the overall efficiency.

Olcay Akman | Joshua W. Hallam | O. Akman | Joshua W. Hallam

[1] R. R. Hocking. The analysis and selection of variables in linear regression , 1976 .

[2] N. Mantel. Why Stepdown Procedures in Variable Selection , 1970 .

[3] John H. Holland,et al. Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[4] A. Propst. Think and explain with statistics , 1988 .

[5] H. Akaike,et al. Information Theory and an Extension of the Maximum Likelihood Principle , 1973 .

[6] R. R. Hocking. Developments in Linear Regression Methodology: 1959–l982 , 1983 .

[7] David E. Goldberg,et al. Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[8] Goldberg,et al. Genetic algorithms , 1993, Robust Control Systems with Genetic Algorithms.

[9] R. Weischedel,et al. Optimal Subset Selection: Multiple Regression, Interdependence and Optimal Network Algorithms , 1974 .

[10] Hamparsum Bozdogan,et al. Statistical Data Mining and Knowledge Discovery , 2004 .

[11] R. A. Fisher,et al. The Genetical Theory of Natural Selection , 1931 .

[12] R. R. Hocking. Developments in linear regression methodology: 1959-1982 , 1983 .

[13] L. Moses,et al. Think and Explain with Statistics , 1988 .