
We develop algorithms to jointly learn a global reward function and agent-specific discount factors from expert demonstrations with different planning horizons.


coremsg

learning-reward-functions-and-discount-factors-from-experts-with-multiple-planning-horizons


Learning Reward Functions and Discount Factors from Experts with Multiple Planning Horizons


title_rewrite


The core message of this paper is that a Bayesian approach to model-based inverse reinforcement learning (BM-IRL) can lead to robust policies by simultaneously estimating the expert's reward function and their internal model of the environment dynamics. This is achieved by incorporating a prior that encodes the accuracy of the expert's dynamics model, which encourages the learner to plan against the worst-case dynamics outside the offline data distribution.


a-bayesian-approach-to-robust-inverse-reinforcement-learning-for-high-dimensional-continuous-control


A Bayesian Approach to Robust Inverse Reinforcement Learning for High-Dimensional Continuous Control
