Semi-Supervised Hierarchical Multi-Label Classifier Based on Local Information

TL;DR

This paper introduces SSHMC-BLI, a semi-supervised hierarchical multi-label classifier that leverages local information and unlabeled data to improve classification performance in complex DAG hierarchies, demonstrated on genomics datasets.

Contribution

The paper presents a novel semi-supervised method for hierarchical multi-label classification that effectively utilizes unlabeled data in DAG hierarchies, improving over purely supervised approaches.

Findings

Unlabeled data improves classification accuracy significantly.

Method achieves statistically significant gains on genomics datasets.

Applicable to various hierarchical classification problems.

Abstract

Scarcity of labeled data is a common problem in supervised classification, since hand-labeling can be time consuming, expensive or hard to label; on the other hand, large amounts of unlabeled information can be found. The problem of scarcity of labeled data is even more notorious in hierarchical classification, because the data of a node is split among its children, which results in few instances associated to the deepest nodes of the hierarchy. In this work it is proposed the semi-supervised hierarchical multi-label classifier based on local information (SSHMC-BLI) which can be trained with labeled and unlabeled data to perform hierarchical classification tasks. The method can be applied to any type of hierarchical problem, here we focus on the most difficult case: hierarchies of DAG type, where the instances can be associated to multiple paths of labels which can finish in an internal node. SSHMC-BLI builds pseudo-labels for each unlabeled instance from the paths of labels of its labeled neighbors, while it considers whether the unlabeled instance is similar to its neighbors. Experiments on 12 challenging datasets from functional genomics show that making use of unlabeled along with labeled data can help to improve the performance of a supervised hierarchical classifier trained only on labeled data, even with statistical significance.