A Method for Speeding Up Value Iteration in Partially Observable Markov Decision Processes
This addresses a computational bottleneck for researchers and practitioners using POMDPs, though it is incremental as it adapts an existing method from fully observable MDPs.
The paper tackles the slow convergence of value iteration in partially observable Markov decision processes (POMDPs) by introducing a technique that speeds up incremental pruning, making it run several orders of magnitude faster in experiments.
We present a technique for speeding up the convergence of value iteration for partially observable Markov decisions processes (POMDPs). The underlying idea is similar to that behind modified policy iteration for fully observable Markov decision processes (MDPs). The technique can be easily incorporated into any existing POMDP value iteration algorithms. Experiments have been conducted on several test problems with one POMDP value iteration algorithm called incremental pruning. We find that the technique can make incremental pruning run several orders of magnitude faster.