Redesigning a multi-objective metaheuristic for the Support Vector Machine with feature selection

J. Alcaraz Soria

The Support Vector Machine is a well-known technique used in supervised
classification. To introduce the selection of features has several benefits but adds
complexity and makes the problem harder. Given that the SVM problem is a multiobjective
problem, obtaining the Pareto front gives the decision maker a wide variety of
solutions where to choose. The only metaheuristic that has been developed to solve
the problem and give an approximation of such a front is a NSGA-II based technique.
However, the design of such technique presents some limitations that are analyzed in
this paper. We present a new metaheuristic that has been completely redesigned in
order to overcome those drawbacks. We compare both techniques through an
extensive computational experiment that demonstrates the superior efficiency of the
new technique.

Palabras clave: Metaheuristics, Support Vector Machine, Feature Selection

Programado

GT10.HEUR1 Sesión Invitada
7 de noviembre de 2023  16:50
HC3: Sala Canónigos 3

Avances en la mejora de procesos en un problema del sector de laminados de acero

Ó. Soto-Sánchez, M. Sierra-Paradinas, M. Gallego, F. J. Martín Campo, F. Gortázar, A. Alonso-Ayuso

Problemas de rutas bicriterio para la circulación segura en vehículos no motorizados en Madrid

C. Simon de Blas, D. Gomez Gonzalez, T. Klinge, J. Cano Cancela