Domain-Splitting Generalized Nogoods from Restarts

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


The use of restarts techniques associated with learning nogoods in solving Constraint Satisfaction Problems (CSPs) is starting to be considered of major importance for backtrack search algorithms. Recent developments show how to learn nogoods from restarts and that those nogoods are essential when using restarts. Using a backtracking search algorithm, with 2-way branching, generalized nogoods are learned from the last branch of the search tree, immediately before the restart occurs. In this paper we further generalized the learned nogoods but now using domain-splitting branching and set branching. We believe that the use of restarts and learning of domain-splitting generalized nogoods will improve backtrack search algorithms for certain classes of problems.
Original languageUnknown
Title of host publicationLNAI
Publication statusPublished - 1 Jan 2011
EventPortuguese Conference on Artificial Intelligence (EPIA) -
Duration: 1 Jan 2011 → …


ConferencePortuguese Conference on Artificial Intelligence (EPIA)
Period1/01/11 → …

Cite this