Název:
LP relaxations and pruning for characteristic imsets
Autoři:
Studený, Milan Typ dokumentu: Výzkumné zprávy
Rok:
2012
Jazyk:
eng
Edice: Research Report, svazek: 2323
Abstrakt: The geometric approach to learning BN structure is to represent it by a certain vector; a suitable such zero-one vector is the characteristic imset, which allows to reformulate the task of finding global maximum of a score over BN structures as an integer linear programming problem. The main contribution of this report is an LP relaxation of the corresponding polytope, that is, a polyhedral description of the domain of the respective integer linear programming problem.
Klíčová slova:
integer linear programming; learning Bayesian network structure; quality criterion Číslo projektu: GA201/08/0539 (CEP) Poskytovatel projektu: GA ČR