Structural-EM for Learning PDG Models from Incomplete Data
Probabilistic Decision Graphs (PDGs) are a class of graphical models that can naturally encode some context specific independencies that cannot always be efficiently captured by other popular models, such as Bayesian Networks. Furthermore, inference can be carried out efficiently over a PDG, in time...
| Main Authors: | , , |
|---|---|
| Format: | info:eu-repo/semantics/report |
| Jezik: | English |
| Izdano: |
2012
|
| Online dostop: | http://hdl.handle.net/10835/1551 |
| _version_ | 1789406619766882304 |
|---|---|
| author | Nielsen, Jens D. Rumí, Rafael Salmerón Cerdán, Antonio |
| author_facet | Nielsen, Jens D. Rumí, Rafael Salmerón Cerdán, Antonio |
| author_sort | Nielsen, Jens D. |
| collection | DSpace |
| description | Probabilistic Decision Graphs (PDGs) are a class of graphical models that can naturally encode some context specific independencies that cannot always be efficiently captured by other popular models, such as Bayesian Networks. Furthermore, inference can be carried out efficiently over a PDG, in time linear in the size of the model. The problem of learning PDGs from data has been studied in the literature, but only for the case of complete data. In this paper we propose an algorithm for learning PDGs in the presence of missing data. The proposed method is based on the EM algorithm for estimating the structure of the model as well as the parameters. We test our proposal on artificially generated data with different rates of missing cells, showing a reasonable performance. |
| format | info:eu-repo/semantics/report |
| id | oai:repositorio.ual.es:10835-1551 |
| institution | Universidad de Cuenca |
| language | English |
| publishDate | 2012 |
| record_format | dspace |
| spelling | oai:repositorio.ual.es:10835-15512023-04-12T19:40:15Z Structural-EM for Learning PDG Models from Incomplete Data Nielsen, Jens D. Rumí, Rafael Salmerón Cerdán, Antonio Probabilistic Decision Graphs (PDGs) are a class of graphical models that can naturally encode some context specific independencies that cannot always be efficiently captured by other popular models, such as Bayesian Networks. Furthermore, inference can be carried out efficiently over a PDG, in time linear in the size of the model. The problem of learning PDGs from data has been studied in the literature, but only for the case of complete data. In this paper we propose an algorithm for learning PDGs in the presence of missing data. The proposed method is based on the EM algorithm for estimating the structure of the model as well as the parameters. We test our proposal on artificially generated data with different rates of missing cells, showing a reasonable performance. 2012-05-28T09:46:58Z 2012-05-28T09:46:58Z 2008 info:eu-repo/semantics/report Proceedings of the Fourth European Workshop on Probabilistic Graphical Models.Pages 217--224. http://hdl.handle.net/10835/1551 en info:eu-repo/semantics/openAccess Fourth European Workshop on Probabilistic Graphical Models |
| spellingShingle | Nielsen, Jens D. Rumí, Rafael Salmerón Cerdán, Antonio Structural-EM for Learning PDG Models from Incomplete Data |
| title | Structural-EM for Learning PDG Models from Incomplete Data |
| title_full | Structural-EM for Learning PDG Models from Incomplete Data |
| title_fullStr | Structural-EM for Learning PDG Models from Incomplete Data |
| title_full_unstemmed | Structural-EM for Learning PDG Models from Incomplete Data |
| title_short | Structural-EM for Learning PDG Models from Incomplete Data |
| title_sort | structural-em for learning pdg models from incomplete data |
| url | http://hdl.handle.net/10835/1551 |
| work_keys_str_mv | AT nielsenjensd structuralemforlearningpdgmodelsfromincompletedata AT rumirafael structuralemforlearningpdgmodelsfromincompletedata AT salmeroncerdanantonio structuralemforlearningpdgmodelsfromincompletedata |