Machine Teaching for Inverse Reinforcement Learning: Algorithms and Applications
This work addresses the efficiency of teaching in IRL for researchers and practitioners, but it is incremental as it builds on prior machine teaching and IRL frameworks.
The paper tackles the problem of determining the minimum number of demonstrations needed to teach a specific sequential decision-making task in inverse reinforcement learning (IRL), by formalizing it as a machine teaching problem and showing a reduction to the set cover problem for an efficient approximation algorithm. It applies this algorithm to provide a lower bound on queries for active IRL and develop a more efficient IRL method from informative demonstrations.
Inverse reinforcement learning (IRL) infers a reward function from demonstrations, allowing for policy improvement and generalization. However, despite much recent interest in IRL, little work has been done to understand the minimum set of demonstrations needed to teach a specific sequential decision-making task. We formalize the problem of finding maximally informative demonstrations for IRL as a machine teaching problem where the goal is to find the minimum number of demonstrations needed to specify the reward equivalence class of the demonstrator. We extend previous work on algorithmic teaching for sequential decision-making tasks by showing a reduction to the set cover problem which enables an efficient approximation algorithm for determining the set of maximally-informative demonstrations. We apply our proposed machine teaching algorithm to two novel applications: providing a lower bound on the number of queries needed to learn a policy using active IRL and developing a novel IRL algorithm that can learn more efficiently from informative demonstrations than a standard IRL approach.