Risk-Sensitive Policy with Distributional Reinforcement Learning
Open Access
- 30 June 2023
- journal article
- research article
- Published by MDPI AG in Algorithms
- Vol. 16 (7) , 325
Abstract
Classical reinforcement learning (RL) techniques are generally concerned with the design of decision-making policies driven by the maximisation of the expected outcome. Nevertheless, this approach does not take into consideration the potential risk associated with the actions taken, which may be critical in certain applications. To address that issue, the present research work introduces a novel methodology based on distributional RL to derive sequential decision-making policies that are sensitive to the risk, the latter being modelled by the tail of the return probability distribution. The core idea is to replace the Q function generally standing at the core of learning schemes in RL by another function, taking into account both the expected return and the risk. Named the risk-based utility function U, it can be extracted from the random return distribution Z naturally learnt by any distributional RL algorithm. This enables the spanning of the complete potential trade-off between risk minimisation and expected return maximisation, in contrast to fully risk-averse methodologies. Fundamentally, this research yields a truly practical and accessible solution for learning risk-sensitive policies with minimal modification to the distributional RL algorithm, with an emphasis on the interpretability of the resulting decision-making process.Keywords
All Related Versions
Funding Information
- Thibaut Théate
This publication has 12 references indexed in Scilit:
- Safety-constrained reinforcement learning with a distributional safety criticMachine Learning, 2022
- A Survey of Deep RL and IL for Autonomous Driving Policy LearningIEEE Transactions on Intelligent Transportation Systems, 2021
- An application of deep reinforcement learning to algorithmic tradingExpert Systems with Applications, 2021
- Challenges of real-world reinforcement learning: definitions, benchmarks and analysisMachine Learning, 2021
- Recovery RL: Safe Reinforcement Learning With Learned Recovery ZonesIEEE Robotics and Automation Letters, 2021
- Guidelines for reinforcement learning in healthcareNature Medicine, 2019
- Optimizing the CVaR via SamplingProceedings of the AAAI Conference on Artificial Intelligence, 2015
- Risk-Sensitive Reinforcement LearningNeural Computation, 2014
- Conditional Value-at-Risk for General Loss DistributionsSSRN Electronic Journal, 2001
- Q-learningMachine Learning, 1992