Regret-based Reward Elicitation for Markov Decision Processes

Understand

The specification of aMarkov decision process (MDP) can be difficult.

• Reward function specification is especially problematic; in practice, it is often cognitively complex and time-consuming for users to precisely specify rewards.
• This work casts the problem of specifying rewards as one of preference elicitation and aims to minimize the degree of precision with which a reward function must be specified while still allowing optimal or near-optimal policies to be produced.
• We first discuss how robust policies can be computed for MDPs given only partial reward information using the minimax regret criterion.