![[Artificial Intelligence and Statistics Logo]](/image.pl?url=http%3a%2f%2fproceedings.mlr.press%2fv33%2fassets%2fimages%2flogo-pmlr.svg&f=jpg&w=240)
[edit]
Active Learning for Undirected Graphical Model Selection
Divyanshu Vats, Robert Nowak, Richard BaraniukProceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, PMLR 33:958-967, 2014.
Abstract
This paper studies graphical model selection, i.e., the problem of estimating a graph of statistical relationships among a collection of random variables. Conventional graphical model selection algorithms are passive, i.e., they require all the measurements to have been collected before processing begins. We propose an active learning algorithm that uses junction tree representations to adapt future measurements based on the information gathered from prior measurements. We prove that, under certain conditions, our active learning algorithm requires fewer scalar measurements than any passive algorithm to reliably estimate a graph. A range of numerical results validate our theory and demonstrates the benefits of active learning.
Cite this Paper
BibTeX
@InProceedings{pmlr-v33-vats14b, title = {{Active Learning for Undirected Graphical Model Selection}}, author = {Vats, Divyanshu and Nowak, Robert and Baraniuk, Richard}, booktitle = {Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics}, pages = {958--967}, year = {2014}, editor = {Kaski, Samuel and Corander, Jukka}, volume = {33}, series = {Proceedings of Machine Learning Research}, address = {Reykjavik, Iceland}, month = {22--25 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v33/vats14b.pdf}, url = {https://proceedings.mlr.press/v33/vats14b.html}, abstract = {This paper studies graphical model selection, i.e., the problem of estimating a graph of statistical relationships among a collection of random variables. Conventional graphical model selection algorithms are passive, i.e., they require all the measurements to have been collected before processing begins. We propose an active learning algorithm that uses junction tree representations to adapt future measurements based on the information gathered from prior measurements. We prove that, under certain conditions, our active learning algorithm requires fewer scalar measurements than any passive algorithm to reliably estimate a graph. A range of numerical results validate our theory and demonstrates the benefits of active learning.}} Endnote
%0 Conference Paper%T Active Learning for Undirected Graphical Model Selection%A Divyanshu Vats%A Robert Nowak%A Richard Baraniuk%B Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics%C Proceedings of Machine Learning Research%D 2014%E Samuel Kaski%E Jukka Corander%F pmlr-v33-vats14b%I PMLR%P 958--967%U https://proceedings.mlr.press/v33/vats14b.html%V 33%X This paper studies graphical model selection, i.e., the problem of estimating a graph of statistical relationships among a collection of random variables. Conventional graphical model selection algorithms are passive, i.e., they require all the measurements to have been collected before processing begins. We propose an active learning algorithm that uses junction tree representations to adapt future measurements based on the information gathered from prior measurements. We prove that, under certain conditions, our active learning algorithm requires fewer scalar measurements than any passive algorithm to reliably estimate a graph. A range of numerical results validate our theory and demonstrates the benefits of active learning. RIS
TY - CPAPERTI - Active Learning for Undirected Graphical Model SelectionAU - Divyanshu VatsAU - Robert NowakAU - Richard BaraniukBT - Proceedings of the Seventeenth International Conference on Artificial Intelligence and StatisticsDA - 2014/04/02ED - Samuel KaskiED - Jukka CoranderID - pmlr-v33-vats14bPB - PMLRDP - Proceedings of Machine Learning ResearchVL - 33SP - 958EP - 967L1 - http://proceedings.mlr.press/v33/vats14b.pdfUR - https://proceedings.mlr.press/v33/vats14b.htmlAB - This paper studies graphical model selection, i.e., the problem of estimating a graph of statistical relationships among a collection of random variables. Conventional graphical model selection algorithms are passive, i.e., they require all the measurements to have been collected before processing begins. We propose an active learning algorithm that uses junction tree representations to adapt future measurements based on the information gathered from prior measurements. We prove that, under certain conditions, our active learning algorithm requires fewer scalar measurements than any passive algorithm to reliably estimate a graph. A range of numerical results validate our theory and demonstrates the benefits of active learning.ER - APA
Vats, D., Nowak, R. & Baraniuk, R.. (2014). Active Learning for Undirected Graphical Model Selection.Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, inProceedings of Machine Learning Research 33:958-967 Available from https://proceedings.mlr.press/v33/vats14b.html.