A Comparison of Reward Functions in Q-Learning Applied to a Cart Position Problem
Published in arXiv Preprint, 2021
Recommended citation: Mukherjee, A. (2021). "A comparison of reward functions in q-learning applied to a cart position problem." arXiv preprint arXiv:2105.11617 https://arxiv.org/abs/2105.11617
In this paper, we explore three reward functions in the cart position problem. This paper concludes that a sparse reward function gives the best results.
Citation: Mukherjee, A. (2021). “A comparison of reward functions in q-learning applied to a cart position problem”. arXiv preprint arXiv:2105.11617