Meta-Learning Transferable Active Learning Policies by Deep Reinforcement Learning

TL;DR

This paper introduces a meta-learning approach using deep reinforcement learning to automatically discover active learning policies that generalize across diverse datasets, reducing annotation costs.

Contribution

It presents a novel method to learn active learning strategies as neural networks trained via reinforcement learning, enabling dataset-agnostic and transferable AL policies.

Findings

Learned AL policies outperform traditional heuristics.

Policies generalize well across heterogeneous datasets.

Reinforcement learning effectively optimizes AL decision-making.

Abstract

Active learning (AL) aims to enable training high performance classifiers with low annotation cost by predicting which subset of unlabelled instances would be most beneficial to label. The importance of AL has motivated extensive research, proposing a wide variety of manually designed AL algorithms with diverse theoretical and intuitive motivations. In contrast to this body of research, we propose to treat active learning algorithm design as a meta-learning problem and learn the best criterion from data. We model an active learning algorithm as a deep neural network that inputs the base learner state and the unlabelled point set and predicts the best point to annotate next. Training this active query policy network with reinforcement learning, produces the best non-myopic policy for a given dataset. The key challenge in achieving a general solution to AL then becomes that of learner generalisation, particularly across heterogeneous datasets. We propose a multi-task dataset-embedding approach that allows dataset-agnostic active learners to be trained. Our evaluation shows that AL algorithms trained in this way can directly generalise across diverse problems.