Probabilistic reasoning in intelligent systems: networks of plausible inference
Probabilistic reasoning in intelligent systems: networks of plausible inference
Chaff: engineering an efficient SAT solver
Proceedings of the 38th annual Design Automation Conference
An Examination of Probabilistic Value-Ordering Heuristics
AI '99 Proceedings of the 12th Australian Joint Conference on Artificial Intelligence: Advanced Topics in Artificial Intelligence
Probabilistic Arc Consistency: A Connection between Constraint Reasoning and Probabilistic Reasoning
UAI '00 Proceedings of the 16th Conference on Uncertainty in Artificial Intelligence
| Hi-index | 0.00 |
Nogood stores are frequently used to avoid revisiting states that were previously discovered to be inconsistent. In this paper we consider the usefulness of learned nogoods as a heuristic to guide search. In particular, we look at learning nogoods probabilistically and examine heuristic utility of such nogoods. We define how probabilistic nogoods can be derived from real nogoods and then introduce an approximate implementation. This implementation is used to compare behavior of heuristics using classic nogoods and then probabilistic nogoods on random binary CSPs and QWH problems. Empirical results show improvement in both problem domains over original heuristics.