Granular conditional entropy-based attribute reduction for partially labeled data with proxy labels

TL;DR

This paper introduces a semi-supervised attribute reduction method based on granular conditional entropy for partially labeled data, effectively utilizing prior class distribution information and a heuristic algorithm to improve attribute selection and classification performance.

Contribution

It proposes a novel granular conditional entropy measure and a heuristic algorithm for semi-supervised attribute reduction in partially labeled data, integrating prior class info and information granularity.

Findings

Outperforms some supervised methods on UCI datasets.

Effectively handles partially labeled data with proxy labels.

Accelerates attribute reduction process.

Abstract

Attribute reduction is one of the most important research topics in the theory of rough sets, and many rough sets-based attribute reduction methods have thus been presented. However, most of them are specifically designed for dealing with either labeled data or unlabeled data, while many real-world applications come in the form of partial supervision. In this paper, we propose a rough sets-based semi-supervised attribute reduction method for partially labeled data. Particularly, with the aid of prior class distribution information about data, we first develop a simple yet effective strategy to produce the proxy labels for unlabeled data. Then the concept of information granularity is integrated into the information-theoretic measure, based on which, a novel granular conditional entropy measure is proposed, and its monotonicity is proved in theory. Furthermore, a fast heuristic algorithm is provided to generate the optimal reduct of partially labeled data, which could accelerate the process of attribute reduction by removing irrelevant examples and excluding redundant attributes simultaneously. Extensive experiments conducted on UCI data sets demonstrate that the proposed semi-supervised attribute reduction method is promising and even compares favourably with the supervised methods on labeled data and unlabeled data with true labels in terms of classification performance.