English

Reinforcement Learning with Non-Exponential Discounting

Machine Learning 2022-12-08 v2 Systems and Control Systems and Control Neurons and Cognition Machine Learning

Abstract

Commonly in reinforcement learning (RL), rewards are discounted over time using an exponential function to model time preference, thereby bounding the expected long-term reward. In contrast, in economics and psychology, it has been shown that humans often adopt a hyperbolic discounting scheme, which is optimal when a specific task termination time distribution is assumed. In this work, we propose a theory for continuous-time model-based reinforcement learning generalized to arbitrary discount functions. This formulation covers the case in which there is a non-exponential random termination time. We derive a Hamilton-Jacobi-Bellman (HJB) equation characterizing the optimal policy and describe how it can be solved using a collocation method, which uses deep learning for function approximation. Further, we show how the inverse RL problem can be approached, in which one tries to recover properties of the discount function given decision data. We validate the applicability of our proposed approach on two simulated problems. Our approach opens the way for the analysis of human discounting in sequential decision-making tasks.

Keywords

Cite

@article{arxiv.2209.13413,
  title  = {Reinforcement Learning with Non-Exponential Discounting},
  author = {Matthias Schultheis and Constantin A. Rothkopf and Heinz Koeppl},
  journal= {arXiv preprint arXiv:2209.13413},
  year   = {2022}
}

Comments

22 pages, 3 figures, published at 36th Conference on Neural Information Processing Systems (NeurIPS 2022)

R2 v1 2026-06-28T02:12:05.992Z