Deep Reinforcement Learning with Quantum-inspired Experience Replay
This work addresses the problem of balancing exploration and exploitation in deep reinforcement learning for researchers and practitioners, offering an incremental improvement over existing experience replay mechanisms.
This paper introduces Deep Reinforcement Learning with Quantum-inspired Experience Replay (DRL-QER), a new training paradigm that adaptively selects experiences from a replay buffer based on complexity and replay frequency. DRL-QER outperforms state-of-the-art algorithms like DRL-PER and DCRL on most Atari 2600 games, demonstrating improved training efficiency.
In this paper, a novel training paradigm inspired by quantum computation is proposed for deep reinforcement learning (DRL) with experience replay. In contrast to traditional experience replay mechanism in DRL, the proposed deep reinforcement learning with quantum-inspired experience replay (DRL-QER) adaptively chooses experiences from the replay buffer according to the complexity and the replayed times of each experience (also called transition), to achieve a balance between exploration and exploitation. In DRL-QER, transitions are first formulated in quantum representations, and then the preparation operation and the depreciation operation are performed on the transitions. In this progress, the preparation operation reflects the relationship between the temporal difference errors (TD-errors) and the importance of the experiences, while the depreciation operation is taken into account to ensure the diversity of the transitions. The experimental results on Atari 2600 games show that DRL-QER outperforms state-of-the-art algorithms such as DRL-PER and DCRL on most of these games with improved training efficiency, and is also applicable to such memory-based DRL approaches as double network and dueling network.