Consistency of plug-in confidence sets for classification in semi-supervised learning

TL;DR

This paper introduces a new methodology for classification with reject options in semi-supervised learning, providing exact confidence control, ease of implementation, and strong theoretical and numerical properties.

Contribution

It develops a novel approach for confidence sets in semi-supervised classification that guarantees exact probability control and is practically implementable.

Findings

Ensures exact control of classification confidence levels.

Demonstrates attractive theoretical properties.

Shows strong numerical performance in experiments.

Abstract

Confident prediction is highly relevant in machine learning; for example, in applications such as medical diagnoses, wrong prediction can be fatal. For classification, there already exist procedures that allow to not classify data when the confidence in their prediction is weak. This approach is known as classification with reject option. In the present paper, we provide new methodology for this approach. Predicting a new instance via a confidence set, we ensure an exact control of the probability of classification. Moreover, we show that this methodology is easily implementable and entails attractive theoretical and numerical properties.