Learning How to Self-Learn: Enhancing Self-Training Using Neural Reinforcement Learning
This work addresses the challenge of reducing manual heuristic adjustments in self-training for semi-supervised learning, offering an incremental improvement in automating the process.
The paper tackles the problem of automating instance selection in self-training for semi-supervised learning by proposing a deep reinforcement learning method, resulting in improved tagging performance and stability compared to baseline solutions.
Self-training is a useful strategy for semi-supervised learning, leveraging raw texts for enhancing model performances. Traditional self-training methods depend on heuristics such as model confidence for instance selection, the manual adjustment of which can be expensive. To address these challenges, we propose a deep reinforcement learning method to learn the self-training strategy automatically. Based on neural network representation of sentences, our model automatically learns an optimal policy for instance selection. Experimental results show that our approach outperforms the baseline solutions in terms of better tagging performances and stability.