Epistemic risk-sensitive reinforcement learning
Paper i proceeding, 2020

We develop a framework for risk-sensitive behaviour in reinforcement learning (RL) due to uncertainty about the environment dynamics by leveraging utility-based definitions of risk sensitivity. In this framework, the preference for risk can be tuned by varying the utility function, for which we develop dynamic programming (DP) and policy gradient-based algorithms. The risk-averse behavior is compared with the behavior of risk-neutral policy in environments with epistemic risk.

Författare

Hannes Eriksson

Chalmers, Data- och informationsteknik, Data Science

Zenuity AB

Christos Dimitrakakis

Universitetet i Oslo

Zenuity AB

ESANN 2020 - Proceedings, 28th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning

339-344

28th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2020
Virtual, Online, Belgium,

Ämneskategorier

Filosofi

Transportteknik och logistik

Annan samhällsbyggnadsteknik

Mer information

Senast uppdaterat

2021-04-28