Copula Network Classifiers (CNCs)

Gal Elidan
Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics, PMLR 22:346-354, 2012.

Abstract

The task of classification is of paramount importance and extensive research has been aimed at developing general purpose classifiers that can be used effectively in a variety of domains. Network-based classifiers, such as the tree augmented naive Bayes model, are appealing since they are easily interpretable, can naturally handle missing data, and are often quite effective. Yet, for complex domains with continuous explanatory variables, practical performance is often sub-optimal. To overcome this limitation, we introduce Copula Network Classifiers (CNCs), a model that combines the flexibility of a graph based representation with the modeling power of copulas. As we demonstrate on ten varied continuous real-life datasets, CNCs offer better overall performance than linear and non-linear standard generative models, as well as discriminative RBF and polynomial kernel SVMs. In addition, since no parameter tuning is required, CNCs can be trained dramatically faster than SVMs.

Cite this Paper

BibTeX
@InProceedings{pmlr-v22-elidan12a, title = {Copula Network Classifiers (CNCs)}, author = {Elidan, Gal}, booktitle = {Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics}, pages = {346--354}, year = {2012}, editor = {Lawrence, Neil D. and Girolami, Mark}, volume = {22}, series = {Proceedings of Machine Learning Research}, address = {La Palma, Canary Islands}, month = {21--23 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v22/elidan12a/elidan12a.pdf}, url = {https://proceedings.mlr.press/v22/elidan12a.html}, abstract = {The task of classification is of paramount importance and extensive research has been aimed at developing general purpose classifiers that can be used effectively in a variety of domains. Network-based classifiers, such as the tree augmented naive Bayes model, are appealing since they are easily interpretable, can naturally handle missing data, and are often quite effective. Yet, for complex domains with continuous explanatory variables, practical performance is often sub-optimal. To overcome this limitation, we introduce Copula Network Classifiers (CNCs), a model that combines the flexibility of a graph based representation with the modeling power of copulas. As we demonstrate on ten varied continuous real-life datasets, CNCs offer better overall performance than linear and non-linear standard generative models, as well as discriminative RBF and polynomial kernel SVMs. In addition, since no parameter tuning is required, CNCs can be trained dramatically faster than SVMs.} }
Endnote
%0 Conference Paper %T Copula Network Classifiers (CNCs) %A Gal Elidan %B Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2012 %E Neil D. Lawrence %E Mark Girolami %F pmlr-v22-elidan12a %I PMLR %P 346--354 %U https://proceedings.mlr.press/v22/elidan12a.html %V 22 %X The task of classification is of paramount importance and extensive research has been aimed at developing general purpose classifiers that can be used effectively in a variety of domains. Network-based classifiers, such as the tree augmented naive Bayes model, are appealing since they are easily interpretable, can naturally handle missing data, and are often quite effective. Yet, for complex domains with continuous explanatory variables, practical performance is often sub-optimal. To overcome this limitation, we introduce Copula Network Classifiers (CNCs), a model that combines the flexibility of a graph based representation with the modeling power of copulas. As we demonstrate on ten varied continuous real-life datasets, CNCs offer better overall performance than linear and non-linear standard generative models, as well as discriminative RBF and polynomial kernel SVMs. In addition, since no parameter tuning is required, CNCs can be trained dramatically faster than SVMs.
RIS
TY - CPAPER TI - Copula Network Classifiers (CNCs) AU - Gal Elidan BT - Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics DA - 2012/03/21 ED - Neil D. Lawrence ED - Mark Girolami ID - pmlr-v22-elidan12a PB - PMLR DP - Proceedings of Machine Learning Research VL - 22 SP - 346 EP - 354 L1 - http://proceedings.mlr.press/v22/elidan12a/elidan12a.pdf UR - https://proceedings.mlr.press/v22/elidan12a.html AB - The task of classification is of paramount importance and extensive research has been aimed at developing general purpose classifiers that can be used effectively in a variety of domains. Network-based classifiers, such as the tree augmented naive Bayes model, are appealing since they are easily interpretable, can naturally handle missing data, and are often quite effective. Yet, for complex domains with continuous explanatory variables, practical performance is often sub-optimal. To overcome this limitation, we introduce Copula Network Classifiers (CNCs), a model that combines the flexibility of a graph based representation with the modeling power of copulas. As we demonstrate on ten varied continuous real-life datasets, CNCs offer better overall performance than linear and non-linear standard generative models, as well as discriminative RBF and polynomial kernel SVMs. In addition, since no parameter tuning is required, CNCs can be trained dramatically faster than SVMs. ER -
APA
Elidan, G.. (2012). Copula Network Classifiers (CNCs). Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 22:346-354 Available from https://proceedings.mlr.press/v22/elidan12a.html.

Related Material