Automated Discovery of Polynomials by Inductive Genetic Programming
This paper presents an approach to automated discovery of high-order multivariate polynomials by inductive Genetic Programming (iGP). Evolutionary search is used for learning polynomials represented as non-linear multivariate trees. Optimal search performance is pursued with balancing the statistical bias and the variance of iGP. We reduce the bias by extending the set of basis polynomials for better agreement with the examples. Possible overfitting due to the reduced bias is conteracted by a variance component, implemented as a regularizing factor of the error in an MDL fitness function. Experimental results demonstrate that regularized iGP discovers accurate, parsimonious, and predictive polynomials when trained on practical data mining tasks.
KeywordsGenetic Program Statistical Bias Basis Polynomial Automate Discovery Data Mining Task
- 2.Freitas, A.A.: A Genetic Programming Framework for two Data Mining Tasks: Classification and Generalized Rule Regression. In: Genetic Programming 1997: Proc. of the Second Annual Conference, pp. 96–101. Morgan Kaufmann, San Francisco (1997)Google Scholar
- 3.Gama, J.: Oblique Linear Tree. In: Liu, X., Cohen, P., Berthold, M. (eds.) Advances in Intelligent Data Analysis IDA 1997, pp. 187–198. Springer, Berlin (1997)Google Scholar
- 5.Iba, H., de Garis, H.: Extending Genetic Programming with Recombinative Guidance. In: Advances in Genetic Programming 2, pp. 69–88. The MIT Press, Cambridge (1996)Google Scholar
- 8.Merz, C.J., Murphy, P.M.: UCI Repository of machine learning databases, Irvine, CA: University of California, Dept. of Inf. and Computer Science (1998), http://www.ics.uci.edu/~mlearn/MLRepository.html
- 10.Quinlan, R.: C4.5: Programs for Machine Learning. Morgan Kaufmann, San Mateo (1993)Google Scholar