Inverse Reinforcement Learning with Multiple Planning Horizons
This addresses a challenge in multi-agent IRL for robotics or AI systems, but it is incremental as it builds on existing IRL methods by handling varying horizons.
The paper tackles the inverse reinforcement learning problem when experts have unknown planning horizons, making reward identification harder, and develops algorithms that learn a global reward function with agent-specific discount factors, demonstrating generalizability across domains.
In this work, we study an inverse reinforcement learning (IRL) problem where the experts are planning under a shared reward function but with different, unknown planning horizons. Without the knowledge of discount factors, the reward function has a larger feasible solution set, which makes it harder for existing IRL approaches to identify a reward function. To overcome this challenge, we develop algorithms that can learn a global multi-agent reward function with agent-specific discount factors that reconstruct the expert policies. We characterize the feasible solution space of the reward function and discount factors for both algorithms and demonstrate the generalizability of the learned reward function across multiple domains.