A robust SVM-based approach with feature selection and outliers detection for classification problems

Resumen

In this talk, a new robust classification model is presented. This model is based on support vector machine (SVM) and deals with outliers detection and feature selection simultaneously. The classifier is built considering the ramp loss margin error and it includes a budget constraint to limit the number of selected features. In this model we use the l1-norm, a norm with the sparse property, particularly suitable for feature selection. The search of this classifier is modeled using a mixed-integer formulation with big M parameters. Two different approaches (exact and heuristic) are proposed to solve the model. The ideas of the exact approach are based on the ones presented in Baldomero-Naranjo M. et al. (2020) while the heuristic approach is based on the Adaptative Kernel Search, see Guastaroba et al. (2017). Finally, the efficiency of the proposed classifier on real-life datasets is shown.