SINDy-RL: Interpretable and Efficient Model-Based Reinforcement Learning
Published in arXiv, 2024
We develop unifying methods for incorporating sparse dictionary learning into RL algorithms to accelerate the training process and provide more interpretable representations of the environment dynamics, reward, and policy.
Recommended citation: Zolman, Nicholas, et al. "SINDy-RL: Interpretable and Efficient Model-Based Reinforcement Learning." arXiv preprint arXiv:2403.09110 (2024). https://arxiv.org/abs/2403.09110