Model-Based Planning with Discrete and Continuous Actions
This addresses a bottleneck in model-based RL for researchers and practitioners by enabling efficient planning in discrete and hybrid action spaces, though it is incremental as it builds on existing differentiable planning approaches.
The paper tackles the problem of model-based planning in discrete action spaces, showing that a simple parameterization with input noise enables effective planning via backprop, matching or outperforming model-free RL and discrete methods on gridworld tasks with limited interactions and extending to combined discrete-continuous actions.
Action planning using learned and differentiable forward models of the world is a general approach which has a number of desirable properties, including improved sample complexity over model-free RL methods, reuse of learned models across different tasks, and the ability to perform efficient gradient-based optimization in continuous action spaces. However, this approach does not apply straightforwardly when the action space is discrete. In this work, we show that it is in fact possible to effectively perform planning via backprop in discrete action spaces, using a simple paramaterization of the actions vectors on the simplex combined with input noise when training the forward model. Our experiments show that this approach can match or outperform model-free RL and discrete planning methods on gridworld navigation tasks in terms of performance and/or planning time while using limited environment interactions, and can additionally be used to perform model-based control in a challenging new task where the action space combines discrete and continuous actions. We furthermore propose a policy distillation approach which yields a fast policy network which can be used at inference time, removing the need for an iterative planning procedure.