State-of-the-art probabilistic planners typically apply look- ahead search and reasoning at each step to make a decision. While this approach can enable high-quality decisions, it can be computationally expensive for problems that require fast decision making. In this paper, we investigate the potential for deep learning to replace search by fast reactive policies. We focus on supervised learning of deep reactive policies for probabilistic planning problems described in RDDL. A key challenge is to explore the large design space of network architectures and training methods, which was critical to prior deep learning successes. We investigate a number of choices in this space and conduct experiments across a set of benchmark problems. Our results show that effective deep reactive policies can be learned for many benchmark problems and that leveraging the planning problem description to define the network structure can be beneficial.
Inspired by the success of the satisfiability approach for deterministic planning, we propose a novel framework for on-line stochastic planning, by embedding the idea of hindsight optimization into a reduction to integer linear programming. In contrast to the previous work using reductions or hindsight optimization, our formulation is general purpose by working with domain specifications over factored state and action spaces, and by doing so is also scalable in principle to exponentially large action spaces. Our approach is competitive with state-of-the-art stochastic planners on challenging benchmark problems, and sometimes exceeds their performance especially in large action spaces.
There are notable examples of online search improving over hand-coded or learned policies (e.g. AlphaZero) for sequential decision making. It is not clear, however, whether or not policy improvement is guaranteed for many of these approaches, even when given a perfect leaf evaluation function and transition model. Indeed, simple counterexamples show that seemingly reasonable online search procedures can hurt performance compared to the original policy. To address this issue, we introduce the choice function framework for analyzing online search procedures for policy improvement. A choice function specifies the actions to be considered at every node of a search tree, with all other actions being pruned. Our main contribution is to give sufficient conditions for stationary and non-stationary choice functions to guarantee that the value achieved by online search is no worse than the original policy. In addition, we describe a general parametric class of choice functions that satisfy those conditions and present an illustrative use case of the empirical utility of the framework.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.