Title: Minimising the delta test for variable selection in regression problems

Authors: Alberto Guillen, Dusan Sovilj, Amaury Lendasse, Fernando Mateo, Ignacio Rojas

Addresses: Department of Informatics, University of Jaen, Campus Las Lagunillas, Edif A3, 23071 Jaen, Spain. ' Department of Information and Computer Science, Helsinki University of Technology, P.O. Box 5400, FI-02015 HUT, Finland. ' Department of Information and Computer Science, Helsinki University of Technology, P.O. Box 5400, FI-02015 HUT, Finland. ' Institute of Applications of Information Technology and Advanced Communications, Polytechnic University of Valencia, Camino de Vera s/n, 46022 Valencia, Spain. ' Department of Computer Architecture and Technology, University of Granada, C/ Periodista Daniel Saucedo s/n, 18071 Granada, Spain.

Abstract: The problem of selecting an adequate set of variables from a given data set of a sampled function becomes crucial by the time of designing the model that will approximate it. Several approaches have been presented in the literature although recent studies showed how the delta test is a powerful tool to determine if a subset of variables is correct. This paper presents new methodologies based on the delta test such as tabu search, genetic algorithms and the hybridisation of them, to determine a subset of variables which is representative of a function. The paper considers as well the scaling problem where a relevance value is assigned to each variable. The new algorithms were adapted to be run in parallel architectures so better performances could be obtained in a small amount of time, presenting great robustness and scalability.

Keywords: variable selection; delta test; forward-backward search; FBS; tabu search; genetic algorithms; GAs; hybrid algorithms; parallel architectures; regression problems.

DOI: 10.1504/IJHPSA.2008.024211

International Journal of High Performance Systems Architecture, 2008 Vol.1 No.4, pp.269 - 281

Published online: 29 Mar 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article