Probabilistic reasoning in intelligent systems: networks of plausible inference
Probabilistic reasoning in intelligent systems: networks of plausible inference
Survey propagation: An algorithm for satisfiability
Random Structures & Algorithms
Backbones and backdoors in satisfiability
AAAI'05 Proceedings of the 20th national conference on Artificial intelligence - Volume 3
Using expectation maximization to find likely assignments for solving CSP's
AAAI'07 Proceedings of the 22nd national conference on Artificial intelligence - Volume 1
Factor graphs and the sum-product algorithm
IEEE Transactions on Information Theory
Deriving Information from Sampling and Diving
Fundamenta Informaticae - RCRA 2009 Experimental Evaluation of Algorithms for Solving Problems with Combinatorial Explosion
Counting-based search: branching heuristics for constraint satisfaction problems
Journal of Artificial Intelligence Research
Hi-index | 0.00 |
Accurately estimating the distribution of solutions to a problem, should such solutions exist, provides efficient search heuristics. The purpose of this paper is to propose new ways of computing such estimates, with different degrees of accuracy and complexity.We build on the Expectation-Maximization Belief-Propagation (EMPB) framework proposed by Hsu et al. to solve Constraint Satisfaction Problems (CSPs). We propose two general approaches within the EMBP framework: we firstly derive update rules at the constraint level while enforcing domain consistency and then derive update rules globally, at the problem level. The contribution of this paper is two-fold: first, we derive new generic update rules suited to tackle any CSP; second, we propose an efficient EMBP-inspired approach, thereby improving this method and making it competitive with the state of the art. We evaluate these approaches experimentally and demonstrate their effectiveness.