Combining Supervised Learning and Reinforcement Learning for Multi-Label Classification Tasks with Partial Labels

TL;DR

This paper introduces MLPAC, an RL-based framework that combines supervised and reinforcement learning to improve multi-label classification with partial labels, addressing annotation challenges in complex tasks.

Contribution

The paper proposes MLPAC, a novel RL-based framework that effectively handles multi-label classification with partial labels, demonstrating improved generalization across various tasks.

Findings

MLPAC outperforms existing methods in document-level relation extraction.

MLPAC achieves higher accuracy in multi-label image classification.

MLPAC demonstrates robustness in binary PU learning scenarios.

Abstract

Traditional supervised learning heavily relies on human-annotated datasets, especially in data-hungry neural approaches. However, various tasks, especially multi-label tasks like document-level relation extraction, pose challenges in fully manual annotation due to the specific domain knowledge and large class sets. Therefore, we address the multi-label positive-unlabelled learning (MLPUL) problem, where only a subset of positive classes is annotated. We propose Mixture Learner for Partially Annotated Classification (MLPAC), an RL-based framework combining the exploration ability of reinforcement learning and the exploitation ability of supervised learning. Experimental results across various tasks, including document-level relation extraction, multi-label image classification, and binary PU learning, demonstrate the generalization and effectiveness of our framework.