Effective input variable selection for function approximation

Herrera, L. J.;Pomares, H.;Rojas, I.;Verleysen, Michel;Guilen, A.
(2006) 16th International Conference on Artificial Neural Networks (ICANN 2006) — Location: Athens (Greece) (10.September.2006)

Files

pdfdocument.pdf
  • Restricted Access
  • Adobe PDF
  • 489.62 KB
Request a copy

Details

Authors
  • Herrera, L. J.University of Granada
    Author
  • Pomares, H.University of Granada
    Author
  • Rojas, I.University of Granada
    Author
  • Verleysen, MichelUCLouvain
    Author
  • Guilen, A.University of Granada
    Author
Abstract
Input variable selection is a key preprocess step in any I/O modelling problem. Normally, better generalization performance is obtained when unneeded parameters coming from irrelevant or redundant variables are eliminated. Information theory provides a robust theoretical framework for performing input variable selection thanks to the concept of mutual information. Nevertheless, for continuous variables, it is usually a more difficult task to determine the mutual information between the input variables and the output variable than for classification problems. This paper presents a modified approach for variable selection for continuous variables adapted from a previous approach for classification problems, making use of a mutual information estimator based on the k-nearest neighbors.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Herrera, L. J., Pomares, H., Rojas, I., Verleysen, M., & Guilen, A. (2006). Effective input variable selection for function approximation. Lecture Notes in Computer Science, 4131, 41-50. https://doi.org/10.1007/11840817_5 (Original work published 2006)