Handling uncertainties in SVM classification

TL;DR

This paper introduces a novel SVM-based approach that effectively handles uncertain target data, including qualitative labels and quantitative probability estimates, improving classification and probability prediction accuracy.

Contribution

It proposes a new SVM-inspired formulation incorporating class label and probability estimates, enabling better handling of uncertainties in classification tasks.

Findings

Outperforms regular SVM in probability prediction accuracy

Provides a dual quadratic formulation with kernel support

Enhances classification performance with uncertain data

Abstract

This paper addresses the pattern classification problem arising when available target data include some uncertainty information. Target data considered here is either qualitative (a class label) or quantitative (an estimation of the posterior probability). Our main contribution is a SVM inspired formulation of this problem allowing to take into account class label through a hinge loss as well as probability estimates using epsilon-insensitive cost function together with a minimum norm (maximum margin) objective. This formulation shows a dual form leading to a quadratic problem and allows the use of a representer theorem and associated kernel. The solution provided can be used for both decision and posterior probability estimation. Based on empirical evidence our method outperforms regular SVM in terms of probability predictions and classification performances.