177 research outputs found
On-Line Parameter Tuning for Monte-Carlo Tree Search in General Game Playing
Full text linkMany enhancements have been proposed for Monte-Carlo Tree Search (MCTS). Some of them have been applied successfully in the context of General Game Playing (GGP). MCTS and its enhancements are usually controlled by multiple parameters that require extensive and time-consuming computation to be tuned in advance. Moreover, in GGP optimal parameter values may vary depending on the considered game. This paper proposes a method to automatically tune search-control parameters on-line for GGP. This method considers the tuning problem as a Combinatorial Multi-Armed Bandit (CMAB). Four strategies designed to deal with CMABs are evaluated for this particular problem. Experiments show that on-line tuning in GGP almost reaches the same performance as off-line tuning. It can be considered as a valid alternative for domains where off-line parameter tuning is costly or infeasible
Parallel hybrid SAT solving using OpenCL
No full textIn the last few decades there have been substantial improvements in approaches for solving the Boolean satisfiability problem. Many of these consisted in elaborating on existing algorithms, both on the side of complete solvers as in the area of incomplete solvers. Besides the improvements to existing solving methods, however, recent evolutions in SAT solving take the form of combining several solvers into one, resulting in parallel solvers and so-called hybrid solvers. Our goal is to combine both approaches, by presenting a parallel hybrid solver. The parallelism exists on two levels: we run a complete solver on the CPU concurrently with an incomplete solver on the GPU, where the latter in turn consists of a massively parallel local search algorithm. We implemented our approach using the OpenCL framework, and present preliminary experimental results.status: Publishe
- …
