Inference of Sparse Networks with Unobserved Variables. Application to Gene Regulatory Networks

Nikolai Slavov
Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, PMLR 9:757-764, 2010.

Abstract

Networks are becoming a unifying framework for modeling complex systems and network inference problems are frequently encountered in many fields. Here, I develop and apply a generative approach to network inference (RCweb) for the case when the network is sparse and the latent (not observed) variables affect the observed ones. From all possible factor analysis (FA) decompositions explaining the variance in the data, RCweb selects the FA decomposition that is consistent with a sparse underlying network. The sparsity constraint is imposed by a novel method that significantly outperforms (in terms of accuracy, robustness to noise, complexity scaling and computational efficiency) methods using $\ell 1$ norm relaxation such as K-SVD and $\ell 1$-based sparse principle component analysis (PCA). Results from simulated models demonstrate that RCweb recovers exactly the model structures for sparsity as low (as non-sparse) as 50% and with ratio of unobserved to observed variables as high as 2. RCweb is robust to noise, with gradual decrease in the parameter ranges as the noise level increases.

Cite this Paper

BibTeX
@InProceedings{pmlr-v9-slavov10a, title = {Inference of Sparse Networks with Unobserved Variables. Application to Gene Regulatory Networks}, author = {Slavov, Nikolai}, booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, pages = {757--764}, year = {2010}, editor = {Teh, Yee Whye and Titterington, Mike}, volume = {9}, series = {Proceedings of Machine Learning Research}, address = {Chia Laguna Resort, Sardinia, Italy}, month = {13--15 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v9/slavov10a/slavov10a.pdf}, url = {https://proceedings.mlr.press/v9/slavov10a.html}, abstract = {Networks are becoming a unifying framework for modeling complex systems and network inference problems are frequently encountered in many fields. Here, I develop and apply a generative approach to network inference (RCweb) for the case when the network is sparse and the latent (not observed) variables affect the observed ones. From all possible factor analysis (FA) decompositions explaining the variance in the data, RCweb selects the FA decomposition that is consistent with a sparse underlying network. The sparsity constraint is imposed by a novel method that significantly outperforms (in terms of accuracy, robustness to noise, complexity scaling and computational efficiency) methods using $\ell 1$ norm relaxation such as K-SVD and $\ell 1$-based sparse principle component analysis (PCA). Results from simulated models demonstrate that RCweb recovers exactly the model structures for sparsity as low (as non-sparse) as 50% and with ratio of unobserved to observed variables as high as 2. RCweb is robust to noise, with gradual decrease in the parameter ranges as the noise level increases.} }
Endnote
%0 Conference Paper %T Inference of Sparse Networks with Unobserved Variables. Application to Gene Regulatory Networks %A Nikolai Slavov %B Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2010 %E Yee Whye Teh %E Mike Titterington %F pmlr-v9-slavov10a %I PMLR %P 757--764 %U https://proceedings.mlr.press/v9/slavov10a.html %V 9 %X Networks are becoming a unifying framework for modeling complex systems and network inference problems are frequently encountered in many fields. Here, I develop and apply a generative approach to network inference (RCweb) for the case when the network is sparse and the latent (not observed) variables affect the observed ones. From all possible factor analysis (FA) decompositions explaining the variance in the data, RCweb selects the FA decomposition that is consistent with a sparse underlying network. The sparsity constraint is imposed by a novel method that significantly outperforms (in terms of accuracy, robustness to noise, complexity scaling and computational efficiency) methods using $\ell 1$ norm relaxation such as K-SVD and $\ell 1$-based sparse principle component analysis (PCA). Results from simulated models demonstrate that RCweb recovers exactly the model structures for sparsity as low (as non-sparse) as 50% and with ratio of unobserved to observed variables as high as 2. RCweb is robust to noise, with gradual decrease in the parameter ranges as the noise level increases.
RIS
TY - CPAPER TI - Inference of Sparse Networks with Unobserved Variables. Application to Gene Regulatory Networks AU - Nikolai Slavov BT - Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics DA - 2010/03/31 ED - Yee Whye Teh ED - Mike Titterington ID - pmlr-v9-slavov10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 9 SP - 757 EP - 764 L1 - http://proceedings.mlr.press/v9/slavov10a/slavov10a.pdf UR - https://proceedings.mlr.press/v9/slavov10a.html AB - Networks are becoming a unifying framework for modeling complex systems and network inference problems are frequently encountered in many fields. Here, I develop and apply a generative approach to network inference (RCweb) for the case when the network is sparse and the latent (not observed) variables affect the observed ones. From all possible factor analysis (FA) decompositions explaining the variance in the data, RCweb selects the FA decomposition that is consistent with a sparse underlying network. The sparsity constraint is imposed by a novel method that significantly outperforms (in terms of accuracy, robustness to noise, complexity scaling and computational efficiency) methods using $\ell 1$ norm relaxation such as K-SVD and $\ell 1$-based sparse principle component analysis (PCA). Results from simulated models demonstrate that RCweb recovers exactly the model structures for sparsity as low (as non-sparse) as 50% and with ratio of unobserved to observed variables as high as 2. RCweb is robust to noise, with gradual decrease in the parameter ranges as the noise level increases. ER -
APA
Slavov, N.. (2010). Inference of Sparse Networks with Unobserved Variables. Application to Gene Regulatory Networks. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 9:757-764 Available from https://proceedings.mlr.press/v9/slavov10a.html.

Related Material