Inductive Transfer for Bayesian Network Structure Learning

Alexandru Niculescu-Mizil, Rich Caruana
Proceedings of ICML Workshop on Unsupervised and Transfer Learning, PMLR 27:167-180, 2012.

Abstract

We study the multi-task Bayesian Network structure learning problem: given data for multiple related problems, learn a Bayesian Network structure for each of them, sharing information among the problems to boost performance. We learn the structures for all the problems simultaneously using a score and search approach that encourages the learned Bayes Net structures to be similar. Encouraging similarity promotes information sharing and prioritizes learning structural features that explain the data from all problems over features that only seem relevant to a single one. This leads to a significant increase in the accuracy of the learned structures, especially when training data is scarce.

Cite this Paper

BibTeX
@InProceedings{pmlr-v27-niculescu12a, title = {Inductive Transfer for Bayesian Network Structure Learning}, author = {Niculescu-Mizil, Alexandru and Caruana, Rich}, booktitle = {Proceedings of ICML Workshop on Unsupervised and Transfer Learning}, pages = {167--180}, year = {2012}, editor = {Guyon, Isabelle and Dror, Gideon and Lemaire, Vincent and Taylor, Graham and Silver, Daniel}, volume = {27}, series = {Proceedings of Machine Learning Research}, address = {Bellevue, Washington, USA}, month = {02 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v27/niculescu12a/niculescu12a.pdf}, url = {https://proceedings.mlr.press/v27/niculescu12a.html}, abstract = {We study the multi-task Bayesian Network structure learning problem: given data for multiple related problems, learn a Bayesian Network structure for each of them, sharing information among the problems to boost performance. We learn the structures for all the problems simultaneously using a score and search approach that encourages the learned Bayes Net structures to be similar. Encouraging similarity promotes information sharing and prioritizes learning structural features that explain the data from all problems over features that only seem relevant to a single one. This leads to a significant increase in the accuracy of the learned structures, especially when training data is scarce.} }
Endnote
%0 Conference Paper %T Inductive Transfer for Bayesian Network Structure Learning %A Alexandru Niculescu-Mizil %A Rich Caruana %B Proceedings of ICML Workshop on Unsupervised and Transfer Learning %C Proceedings of Machine Learning Research %D 2012 %E Isabelle Guyon %E Gideon Dror %E Vincent Lemaire %E Graham Taylor %E Daniel Silver %F pmlr-v27-niculescu12a %I PMLR %P 167--180 %U https://proceedings.mlr.press/v27/niculescu12a.html %V 27 %X We study the multi-task Bayesian Network structure learning problem: given data for multiple related problems, learn a Bayesian Network structure for each of them, sharing information among the problems to boost performance. We learn the structures for all the problems simultaneously using a score and search approach that encourages the learned Bayes Net structures to be similar. Encouraging similarity promotes information sharing and prioritizes learning structural features that explain the data from all problems over features that only seem relevant to a single one. This leads to a significant increase in the accuracy of the learned structures, especially when training data is scarce.
RIS
TY - CPAPER TI - Inductive Transfer for Bayesian Network Structure Learning AU - Alexandru Niculescu-Mizil AU - Rich Caruana BT - Proceedings of ICML Workshop on Unsupervised and Transfer Learning DA - 2012/06/27 ED - Isabelle Guyon ED - Gideon Dror ED - Vincent Lemaire ED - Graham Taylor ED - Daniel Silver ID - pmlr-v27-niculescu12a PB - PMLR DP - Proceedings of Machine Learning Research VL - 27 SP - 167 EP - 180 L1 - http://proceedings.mlr.press/v27/niculescu12a/niculescu12a.pdf UR - https://proceedings.mlr.press/v27/niculescu12a.html AB - We study the multi-task Bayesian Network structure learning problem: given data for multiple related problems, learn a Bayesian Network structure for each of them, sharing information among the problems to boost performance. We learn the structures for all the problems simultaneously using a score and search approach that encourages the learned Bayes Net structures to be similar. Encouraging similarity promotes information sharing and prioritizes learning structural features that explain the data from all problems over features that only seem relevant to a single one. This leads to a significant increase in the accuracy of the learned structures, especially when training data is scarce. ER -
APA
Niculescu-Mizil, A. & Caruana, R.. (2012). Inductive Transfer for Bayesian Network Structure Learning. Proceedings of ICML Workshop on Unsupervised and Transfer Learning, in Proceedings of Machine Learning Research 27:167-180 Available from https://proceedings.mlr.press/v27/niculescu12a.html.

Related Material