This paper is an extended version of the authors' paper12 presented at JCIS 2007. Standard Monte Carlo UCT tree search algorithm is modified and extended to provide an efficient Go capturing problem solver. Experimental results show that this method outperforms traditional tree search methods to...

Monte-Carlo Tree Search (MCTS) is a new best-first search guided by the results of Monte-Carlo simulations. In this article, we introduce two progressive strategies for MCTS, called progressive bias and progressive unpruning. They enable the use of relatively time-expensive heuristic knowledge...

We study connectivity of stones on a Go Board, including influence and heuristics-based soft connectivity and pattern- and search-based hard connectivity. Stone connectivity provides a Go program with an understanding of a fundamental aspect of board configurations. It is essential in positional...

The separation game is different from the connection game, but has some similarities. In the game of Go, it is often useful because it helps to enclose groups and areas. An evaluation function, moves generation functions and a search algorithm for the separation game are described in this paper.