2020-07-09T04:45:42Z
http://oa.upm.es/cgi/oai2
oai:oa.upm.es:54651
2020-01-24T11:01:24Z
7374617475733D707562
7375626A656374733D696E666F726D6174696361
747970653D636F6E666572656E63655F6974656D
Learning Bayesian network classifiers with completed partially directed acyclic graphs
Mihaljevic, Bojan
Bielza Lozoya, María Concepción
Larrañaga Múgica, Pedro María
Computer Science
Most search and score algorithms for learning Bayesian network classifiers from data traverse the space of directed acyclic graphs (DAGs), making arbitrary yet possibly suboptimal arc directionality decisions. This can be remedied by learning in the space of DAG equivalence classes. We provide a number of contributions to existing work along this line. First, we identify the smallest subspace of DAGs that covers all possible class-posterior distributions when data is complete. All the DAGs in this space, which we call minimal class-focused DAGs (MC-DAGs), are such that their every arc is directed towards a child of the class variable. Second, in order to traverse the equivalence classes of MC-DAGs, we adapt the greedy equivalence search (GES) by adding operator validity criteria which ensure GES only visits states within our space. Third, we specify how to efficiently evaluate the discriminative score of a GES operator for MC-DAG in time independent of the number of variables and without converting the completed partially DAG, which represents an equivalence class, into a DAG. The adapted GES perfomed well on real-world data sets.
E.T.S.I de Sistemas Informáticos (UPM)
http://creativecommons.org/licenses/by-nc-nd/3.0/es/
2018
info:eu-repo/semantics/conferenceObject
Presentation at Congress or Conference
Proceedings of Machine Learning Research | International Conference on Probabilistic Graphical Models | 11(09/2018-14/09/2018 | Praga, República Checa
PeerReviewed
application/pdf
eng
http://proceedings.mlr.press/v72/
info:eu-repo/grantAgreement/EC/H2020/785907/EU/Human Brain Project Specific Grant Agreement 2/HBP SGA2
C080020-09
TIN2016-79684-P
S2013/ICE-2845
info:eu-repo/semantics/openAccess
http://oa.upm.es/54651/