Learning Symbolic Persistent Macro-Actions for POMDP Solving Over Time
This addresses the challenge of reducing inference time and improving robustness in POMDP solving for domains requiring interpretable, long-term planning, though it appears incremental as it builds on existing MCTS and LTL methods.
The paper tackled the problem of interpretable decision-making under uncertainty in POMDPs by integrating temporal logical reasoning to generate persistent macro-actions, resulting in significant computational efficiency improvements in benchmark scenarios like Pocman and Rocksample.
This paper proposes an integration of temporal logical reasoning and Partially Observable Markov Decision Processes (POMDPs) to achieve interpretable decision-making under uncertainty with macro-actions. Our method leverages a fragment of Linear Temporal Logic (LTL) based on Event Calculus (EC) to generate \emph{persistent} (i.e., constant) macro-actions, which guide Monte Carlo Tree Search (MCTS)-based POMDP solvers over a time horizon, significantly reducing inference time while ensuring robust performance. Such macro-actions are learnt via Inductive Logic Programming (ILP) from a few traces of execution (belief-action pairs), thus eliminating the need for manually designed heuristics and requiring only the specification of the POMDP transition model. In the Pocman and Rocksample benchmark scenarios, our learned macro-actions demonstrate increased expressiveness and generality when compared to time-independent heuristics, indeed offering substantial computational efficiency improvements.