Open Access
2014 Estimation of high-dimensional partially-observed discrete Markov random fields
Yves F. Atchade
Electron. J. Statist. 8(2): 2242-2263 (2014). DOI: 10.1214/14-EJS946


We consider the problem of estimating the parameters of discrete Markov random fields from partially observed data in a high-dimensional setting. Using a $\ell^{1}$-penalized pseudo-likelihood approach, we fit a misspecified model obtained by ignoring the missing data problem. We derive an estimation error bound that highlights the effect of the misspecification. We report some simulation results that illustrate the theoretical findings.


Download Citation

Yves F. Atchade. "Estimation of high-dimensional partially-observed discrete Markov random fields." Electron. J. Statist. 8 (2) 2242 - 2263, 2014.


Published: 2014
First available in Project Euclid: 31 October 2014

zbMATH: 1302.62206
MathSciNet: MR3273625
Digital Object Identifier: 10.1214/14-EJS946

Primary: 62G20 , 62M40

Keywords: high-dimensional inference , Markov random fields , misspecification , network estimation , penalized likelihood inference , pseudo-likelihood

Rights: Copyright © 2014 The Institute of Mathematical Statistics and the Bernoulli Society

Vol.8 • No. 2 • 2014
Back to Top