Multi-Label Classification with Cutset Networks

Nicola Di Mauro, Antonio Vergari, Floriana Esposito
; Proceedings of the Eighth International Conference on Probabilistic Graphical Models, PMLR 52:147-158, 2016.

Abstract

In this work, we tackle the problem of Multi-Label Classification (MLC) by using Cutset Networks (CNets), weighted probabilistic model trees, recently proposed as \emphtractable probabilistic models for discrete distributions. We employ CNets to perform Most Probable Explanation (MPE) inference exactly and efficiently and we improve a state-of-the-art structure learning algorithm for CNets by explicitly taking advantage of label dependencies. We achieve this by forcing the tree inner nodes to represent only feature variables and by exploiting structural heuristics while learning the leaf models. A thorough experimental evaluation on ten real-world datasets shows how the proposed approach improves several metrics for MLC, proving it to be competitive with problem transformation methods like classifier chains.

Cite this Paper


BibTeX
@InProceedings{pmlr-v52-dimauro16, title = {Multi-Label Classification with Cutset Networks}, author = {Nicola Di Mauro and Antonio Vergari and Floriana Esposito}, booktitle = {Proceedings of the Eighth International Conference on Probabilistic Graphical Models}, pages = {147--158}, year = {2016}, editor = {Alessandro Antonucci and Giorgio Corani and Cassio Polpo Campos}}, volume = {52}, series = {Proceedings of Machine Learning Research}, address = {Lugano, Switzerland}, month = {06--09 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v52/dimauro16.pdf}, url = {http://proceedings.mlr.press/v52/dimauro16.html}, abstract = {In this work, we tackle the problem of Multi-Label Classification (MLC) by using Cutset Networks (CNets), weighted probabilistic model trees, recently proposed as \emphtractable probabilistic models for discrete distributions. We employ CNets to perform Most Probable Explanation (MPE) inference exactly and efficiently and we improve a state-of-the-art structure learning algorithm for CNets by explicitly taking advantage of label dependencies. We achieve this by forcing the tree inner nodes to represent only feature variables and by exploiting structural heuristics while learning the leaf models. A thorough experimental evaluation on ten real-world datasets shows how the proposed approach improves several metrics for MLC, proving it to be competitive with problem transformation methods like classifier chains.} }
Endnote
%0 Conference Paper %T Multi-Label Classification with Cutset Networks %A Nicola Di Mauro %A Antonio Vergari %A Floriana Esposito %B Proceedings of the Eighth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2016 %E Alessandro Antonucci %E Giorgio Corani %E Cassio Polpo Campos} %F pmlr-v52-dimauro16 %I PMLR %J Proceedings of Machine Learning Research %P 147--158 %U http://proceedings.mlr.press %V 52 %W PMLR %X In this work, we tackle the problem of Multi-Label Classification (MLC) by using Cutset Networks (CNets), weighted probabilistic model trees, recently proposed as \emphtractable probabilistic models for discrete distributions. We employ CNets to perform Most Probable Explanation (MPE) inference exactly and efficiently and we improve a state-of-the-art structure learning algorithm for CNets by explicitly taking advantage of label dependencies. We achieve this by forcing the tree inner nodes to represent only feature variables and by exploiting structural heuristics while learning the leaf models. A thorough experimental evaluation on ten real-world datasets shows how the proposed approach improves several metrics for MLC, proving it to be competitive with problem transformation methods like classifier chains.
RIS
TY - CPAPER TI - Multi-Label Classification with Cutset Networks AU - Nicola Di Mauro AU - Antonio Vergari AU - Floriana Esposito BT - Proceedings of the Eighth International Conference on Probabilistic Graphical Models PY - 2016/08/15 DA - 2016/08/15 ED - Alessandro Antonucci ED - Giorgio Corani ED - Cassio Polpo Campos} ID - pmlr-v52-dimauro16 PB - PMLR SP - 147 DP - PMLR EP - 158 L1 - http://proceedings.mlr.press/v52/dimauro16.pdf UR - http://proceedings.mlr.press/v52/dimauro16.html AB - In this work, we tackle the problem of Multi-Label Classification (MLC) by using Cutset Networks (CNets), weighted probabilistic model trees, recently proposed as \emphtractable probabilistic models for discrete distributions. We employ CNets to perform Most Probable Explanation (MPE) inference exactly and efficiently and we improve a state-of-the-art structure learning algorithm for CNets by explicitly taking advantage of label dependencies. We achieve this by forcing the tree inner nodes to represent only feature variables and by exploiting structural heuristics while learning the leaf models. A thorough experimental evaluation on ten real-world datasets shows how the proposed approach improves several metrics for MLC, proving it to be competitive with problem transformation methods like classifier chains. ER -
APA
Di Mauro, N., Vergari, A. & Esposito, F.. (2016). Multi-Label Classification with Cutset Networks. Proceedings of the Eighth International Conference on Probabilistic Graphical Models, in PMLR 52:147-158

Related Material