Learning Structural Changes of Gaussian Graphical Models in Controlled Experiments

Bai Zhang, Yue Wang
Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, PMLR R8:726-733, 2010.

Abstract

Graphical models are widely used in scien- tific and engineering research to represent conditional independence structures between random variables. In many controlled ex- periments, environmental changes or exter- nal stimuli can often alter the conditional dependence between the random variables, and potentially produce significant structural changes in the corresponding graphical mod- els. Therefore, it is of great importance to be able to detect such structural changes from data, so as to gain novel insights into where and how the structural changes take place and help the system adapt to the new environment. Here we report an effec- tive learning strategy to extract structural changes in Gaussian graphical model using $\ell$1-regularization based convex optimization. We discuss the properties of the problem for- mulation and introduce an efficient imple- mentation by the block coordinate descent algorithm. We demonstrate the principle of the approach on a numerical simulation ex- periment, and we then apply the algorithm to the modeling of gene regulatory networks un- der different conditions and obtain promising yet biologically plausible results.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR8-zhang10d, title = {Learning Structural Changes of {G}aussian Graphical Models in Controlled Experiments}, author = {Zhang, Bai and Wang, Yue}, booktitle = {Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence}, pages = {726--733}, year = {2010}, editor = {Grünwald, Peter and Spirtes, Peter}, volume = {R8}, series = {Proceedings of Machine Learning Research}, month = {08--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r8/main/assets/zhang10d/zhang10d.pdf}, url = {https://proceedings.mlr.press/r8/zhang10d.html}, abstract = {Graphical models are widely used in scien- tific and engineering research to represent conditional independence structures between random variables. In many controlled ex- periments, environmental changes or exter- nal stimuli can often alter the conditional dependence between the random variables, and potentially produce significant structural changes in the corresponding graphical mod- els. Therefore, it is of great importance to be able to detect such structural changes from data, so as to gain novel insights into where and how the structural changes take place and help the system adapt to the new environment. Here we report an effec- tive learning strategy to extract structural changes in Gaussian graphical model using $\ell$1-regularization based convex optimization. We discuss the properties of the problem for- mulation and introduce an efficient imple- mentation by the block coordinate descent algorithm. We demonstrate the principle of the approach on a numerical simulation ex- periment, and we then apply the algorithm to the modeling of gene regulatory networks un- der different conditions and obtain promising yet biologically plausible results.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Learning Structural Changes of Gaussian Graphical Models in Controlled Experiments %A Bai Zhang %A Yue Wang %B Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2010 %E Peter Grünwald %E Peter Spirtes %F pmlr-vR8-zhang10d %I PMLR %P 726--733 %U https://proceedings.mlr.press/r8/zhang10d.html %V R8 %X Graphical models are widely used in scien- tific and engineering research to represent conditional independence structures between random variables. In many controlled ex- periments, environmental changes or exter- nal stimuli can often alter the conditional dependence between the random variables, and potentially produce significant structural changes in the corresponding graphical mod- els. Therefore, it is of great importance to be able to detect such structural changes from data, so as to gain novel insights into where and how the structural changes take place and help the system adapt to the new environment. Here we report an effec- tive learning strategy to extract structural changes in Gaussian graphical model using $\ell$1-regularization based convex optimization. We discuss the properties of the problem for- mulation and introduce an efficient imple- mentation by the block coordinate descent algorithm. We demonstrate the principle of the approach on a numerical simulation ex- periment, and we then apply the algorithm to the modeling of gene regulatory networks un- der different conditions and obtain promising yet biologically plausible results. %Z Reissued by PMLR on 04 October 2026.
APA
Zhang, B. & Wang, Y.. (2010). Learning Structural Changes of Gaussian Graphical Models in Controlled Experiments. Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R8:726-733 Available from https://proceedings.mlr.press/r8/zhang10d.html. Reissued by PMLR on 04 October 2026.

Related Material