Variable selection for the multicategory SVM via adaptive sup-norm regularization

TL;DR

This paper introduces adaptive sup-norm regularization for multicategory SVMs, enabling automatic variable selection, improved interpretability, and enhanced accuracy, especially in high-dimensional, low-sample-size scenarios.

Contribution

It proposes novel adaptive sup-norm regularization methods for multicategory SVMs that automate variable selection and improve classifier performance.

Findings

Outperforms existing methods in simulated data.

Effective in high-dimensional, low-sample-size data.

Demonstrated success on real gene expression data.

Abstract

The Support Vector Machine (SVM) is a popular classification paradigm in machine learning and has achieved great success in real applications. However, the standard SVM can not select variables automatically and therefore its solution typically utilizes all the input variables without discrimination. This makes it difficult to identify important predictor variables, which is often one of the primary goals in data analysis. In this paper, we propose two novel types of regularization in the context of the multicategory SVM (MSVM) for simultaneous classification and variable selection. The MSVM generally requires estimation of multiple discriminating functions and applies the argmax rule for prediction. For each individual variable, we propose to characterize its importance by the supnorm of its coefficient vector associated with different functions, and then minimize the MSVM hinge loss function subject to a penalty on the sum of supnorms. To further improve the supnorm penalty, we propose the adaptive regularization, which allows different weights imposed on different variables according to their relative importance. Both types of regularization automate variable selection in the process of building classifiers, and lead to sparse multi-classifiers with enhanced interpretability and improved accuracy, especially for high dimensional low sample size data. One big advantage of the supnorm penalty is its easy implementation via standard linear programming. Several simulated examples and one real gene data analysis demonstrate the outstanding performance of the adaptive supnorm penalty in various data settings.