PERFORMANCE ANALYSIS OF A QUANTUM-CLASSICAL HYBRID REINFORCEMENT LEARNING APPROACH

Evan Mitchell, BISWAJIT BASU, Pabitra Mitra

Keywords: Reinforcement Learning, Quantum Machine Learning, Hybrid Learning, Variational Quantum Circuits

TL;DR: This paper proposes a Qiskit and PyTorch hybrid quantum-classical reinforcement learning model that achieves greater learning curve stability and higher median reward than equivalent classical counterparts in the CartPole-V0 environment.

Abstract: Quantum Machine Learning (QML) is a nascent field of technology that is yet to be fully explored. While previous QML implementations have demonstrated performance efficiency gains over classical benchmarks, it has not been studied in detail whether shallow unentangled quantum circuits can provide the same benefits to reinforcement learning algorithms. Towards this goal, we present a shallow Deep Q-Network (DQN) hybrid quantum-classical Variational Quantum Circuit (VQC) model in the Cartpole-v0 environment that provides an increase in training stability and average reward for any given training run with a simpler unentangled quantum circuit than what is proposed in prior literature.

Submission Number: 202