Variable and Value Ordering for MPE Search
نویسندگان
چکیده
In Bayesian networks, amost probable explanation (MPE) is a most likely instantiation of all network variables given a piece of evidence. Recent work proposed a branch-and-boundsearch algorithm that finds exact solutions to MPE queries, where bounds are computed on a relaxed network obtained by a technique known as node splitting. In this work we study the impact of variable and value ordering on such a search algorithm. We study several heuristics based on the entropies of variables and on the notion of nogoods, and propose a new meta-heuristic that combines their strengths. Experiments indicate that search efficiency is significantly improved, allowing many hard problems to be solved for the first time.
منابع مشابه
Systematic vs. Non-systematic Algorithms for Solving the MPE Task
The paper explores the power of two system atic Branch and Bound search algorithms that exploit partition-based heuristics, BBBT (a new algorithm for which the heuristic information is constructed during search and allows dy namic variable/value ordering) and its predeces sor BBMB (for which the heuristic information is pre-compiled) and compares them against a number of popular local search...
متن کاملWinning the PASCAL 2011 MAP Challenge with Enhanced AND/OR Branch-and-Bound
This paper describes our entry for the MAP/MPE track of the PASCAL 2011 Probabilistic Inference Challenge, which placed first in all three time limit categories, 20 seconds, 20 minutes, and 1 hour. Our baseline is a branch-and-bound algorithm that explores the AND/OR context-minimal search graph of a graphical model guided by a mini-bucket heuristic. Augmented with recent advances that convert ...
متن کاملLook-Ahead Value Ordering for Constraint Satisfaction Problems
Looking ahead during search is often useful when solving constraint satisfaction problems. Previous studies have shown that looking ahead helps by causing dead-ends to occur earlier in the search, and by providing information that is useful for dynamic variable ordering. In this paper, we show that another bene t of looking ahead is a useful domain value ordering heuristic, which we call look-a...
متن کاملAND/OR Branch-and-Bound search for combinatorial optimization in graphical models
We introduce a new generation of depth-first Branch-and-Bound algorithms that explore the AND/OR search tree using static and dynamic variable orderings for solving general constraint optimization problems. The virtue of the AND/OR representation of the search space is that its size may be far smaller than that of a traditional OR representation, which can translate into significant time saving...
متن کاملValue Ordering for Finding All Solutions
In finding all solutions to a constraint satisfaction problem, or proving that there are none, with a search algorithm that backtracks chronologically and forms k-way branches, the order in which the values are assigned is immaterial. However, we show that if the values of a variable are assigned instead via a sequence of binary choice points, and the removal of the value just tried from the do...
متن کامل