Timezone: »

Learner-aware Teaching: Inverse Reinforcement Learning with Preferences and Constraints
Sebastian Tschiatschek · Ahana Ghosh · Luis Haug · Rati Devidze · Adish Singla

Tue Dec 10 10:45 AM -- 12:45 PM (PST) @ East Exhibition Hall B + C #189

Inverse reinforcement learning (IRL) enables an agent to learn complex behavior by observing demonstrations from a (near-)optimal policy. The typical assumption is that the learner's goal is to match the teacher’s demonstrated behavior. In this paper, we consider the setting where the learner has its own preferences that it additionally takes into consideration. These preferences can for example capture behavioral biases, mismatched worldviews, or physical constraints. We study two teaching approaches: learner-agnostic teaching, where the teacher provides demonstrations from an optimal policy ignoring the learner's preferences, and learner-aware teaching, where the teacher accounts for the learner’s preferences. We design learner-aware teaching algorithms and show that significant performance improvements can be achieved over learner-agnostic teaching.

Author Information

Sebastian Tschiatschek (Microsoft Research)
Ahana Ghosh (MPI-SWS)
Luis Haug (ETH Zurich)
Rati Devidze (MPI-SWS)
Adish Singla (MPI-SWS)

More from the Same Authors