Generalized Proof Number Search
نویسندگان
چکیده
We present Generalized Proof Number Search (GPNS), a Proof Number based algorithm able to prove positions in games with multiple outcomes. GPNS is a direct generalization of Proof Number Search (PNS) in the sense that both behave exactly the same way in games with two outcomes. However, GPNS targets a wider class of games. When a game features more than two outcomes, PNS can be used multiple times with different objectives to finally deduce the value of a position. On the contrary, GPNS is called only once to produce the same information. We present experimental results on solving various sizes of the games Connect Four and Woodpush showing that the total number of tree descents of GPNS is much lower than the cumulative number of tree descents of PNS.
منابع مشابه
A Non-clausal Connection Calculus
A non-clausal connection calculus for classical first-order logic is presented that does not require the translation of input formulae into any clausal form. The definition of clauses is generalized, which may now also contain (sub-) matrices. Copying of appropriate (sub-)clauses in a dynamic way, i.e. during the actual proof search, is realized by a generalized extension rule. Thus, the calcul...
متن کاملMonte-Carlo Proof-Number Search for Computer Go
In the last decade, proof-number search and Monte-Carlo methods have successfully been applied to the combinatorial-games domain. Proof-number search is a reliable algorithm. It requires a well defined goal to prove. This can be seen as a disadvantage. In contrast to proof-number search, Monte-Carlo evaluation is a flexible stochastic evaluation for game-tree search. In order to improve the eff...
متن کاملProof - Set Search by Martin Müller
Victor Allis’ proof-number search is a powerful best-first tree search method which can solve games by repeatedly expanding a most-proving node in the game tree. A well-known problem of proof-number search is that it does not account for the effect of transpositions. If the search builds a directed acyclic graph instead of a tree, the same node can be counted more than once, leading to incorrec...
متن کاملProof-Set Search
Victor Allis’ proof-number search is a powerful best-first tree search method which can solve games by repeatedly expanding a most-proving node in the game tree. A well-known problem of proof-number search is that it does not account for the effect of transpositions. If the search builds a directed acyclic graph instead of a tree, the same node can be counted more than once, leading to incorrec...
متن کاملQ-Resolution with Generalized Axioms
Q-resolution is a proof system for quantified Boolean formulas (QBFs) in prenex conjunctive normal form (PCNF) which underlies search-based QBF solvers with clause and cube learning (QCDCL). With the aim to derive and learn stronger clauses and cubes earlier in the search, we generalize the axioms of the Q-resolution calculus resulting in an exponentially more powerful proof system. The general...
متن کامل