Learning safe, constrained policies via imitation learning: Connection to Probabilistic Inference and a Naive Algorithm
This work addresses the need for safe and constrained policy learning in imitation learning, though it appears incremental as it builds on existing connections to probabilistic inference and entropy maximization.
The paper tackles the problem of learning safe, constrained policies via imitation learning by introducing a method that connects to probabilistic inference and uses dual gradient descent for optimization. Experiments show the method learns effective policies that comply with multiple constraints and generalize across different modalities.
This article introduces an imitation learning method for learning maximum entropy policies that comply with constraints demonstrated by expert trajectories executing a task. The formulation of the method takes advantage of results connecting performance to bounds for the KL-divergence between demonstrated and learned policies, and its objective is rigorously justified through a connection to a probabilistic inference framework for reinforcement learning, incorporating the reinforcement learning objective and the objective to abide by constraints in an entropy maximization setting. The proposed algorithm optimizes the learning objective with dual gradient descent, supporting effective and stable training. Experiments show that the proposed method can learn effective policy models for constraints-abiding behaviour, in settings with multiple constraints of different types, accommodating different modalities of demonstrated behaviour, and with abilities to generalize.