TY - GEN
T1 - The nonparanormal SKEPTIC
AU - Liu, Han
AU - Han, Fang
AU - Yuan, Ming
AU - Lafferty, John
AU - Wasserman, Larry
PY - 2012/10/10
Y1 - 2012/10/10
N2 - We propose a semiparametric method we call the nonparanormal SKEPTIC for estimating high dimensional undirected graphical models. The underlying model is the nonparanormal family proposed by Liu et al. (2009). The method exploits nonparametric rank-based correlation coefficient estimators, including Spearman's rho and Kendall's tau. In high dimensional settings, we prove that the nonparanormal SKEPTIC achieves the optimal parametric rate of convergence for both graph and parameter estimation. This result suggests that the nonparanormal graphical model can be a safe replacement for the Gaussian graphical model, even when the data are Gaussian.
AB - We propose a semiparametric method we call the nonparanormal SKEPTIC for estimating high dimensional undirected graphical models. The underlying model is the nonparanormal family proposed by Liu et al. (2009). The method exploits nonparametric rank-based correlation coefficient estimators, including Spearman's rho and Kendall's tau. In high dimensional settings, we prove that the nonparanormal SKEPTIC achieves the optimal parametric rate of convergence for both graph and parameter estimation. This result suggests that the nonparanormal graphical model can be a safe replacement for the Gaussian graphical model, even when the data are Gaussian.
UR - http://www.scopus.com/inward/record.url?scp=84867128960&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84867128960&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84867128960
SN - 9781450312851
T3 - Proceedings of the 29th International Conference on Machine Learning, ICML 2012
SP - 1415
EP - 1422
BT - Proceedings of the 29th International Conference on Machine Learning, ICML 2012
T2 - 29th International Conference on Machine Learning, ICML 2012
Y2 - 26 June 2012 through 1 July 2012
ER -