Positive-Unlabeled Classification under Class Prior Shift and Asymmetric Error

TL;DR

This paper extends positive-unlabeled classification to scenarios with class prior shift and asymmetric errors, proposing two frameworks and demonstrating their effectiveness through experiments.

Contribution

It introduces novel frameworks for PU classification under class prior shift and asymmetric errors, addressing practical limitations of previous assumptions.

Findings

Both frameworks outperform baseline methods in experiments.

The proposed methods effectively handle class prior shift.

Density ratio estimation framework shows competitive results.

Abstract

Bottlenecks of binary classification from positive and unlabeled data (PU classification) are the requirements that given unlabeled patterns are drawn from the test marginal distribution, and the penalty of the false positive error is identical to the false negative error. However, such requirements are often not fulfilled in practice. In this paper, we generalize PU classification to the class prior shift and asymmetric error scenarios. Based on the analysis of the Bayes optimal classifier, we show that given a test class prior, PU classification under class prior shift is equivalent to PU classification with asymmetric error. Then, we propose two different frameworks to handle these problems, namely, a risk minimization framework and density ratio estimation framework. Finally, we demonstrate the effectiveness of the proposed frameworks and compare both frameworks through experiments using benchmark datasets.