The log-linear group-lasso estimator and its asymptotic properties

Abstract

We define the group-lasso estimator for the natural parameters of the exponential families of distributions representing hierarchical log-linear models under multinomial sampling scheme. Such estimator arises as the solution of a convex penalized likelihood optimization problem based on the group-lasso penalty. We illustrate how it is possible to construct an estimator of the underlying log-linear model using the blocks of nonzero coefficients recovered by the group-lasso procedure. We investigate the asymptotic properties of the group-lasso estimator as a model selection method in a double-asymptotic framework, in which both the sample size and the model complexity grow simultaneously. We provide conditions guaranteeing that the group-lasso estimator is model selection consistent, in the sense that, with overwhelming probability as the sample size increases, it correctly identifies all the sets of nonzero interactions among the variables. Provided the sequences of true underlying models is sparse enough, recovery is possible even if the number of cells grows larger than the sample size. Finally, we derive some central limit type of results for the log-linear group-lasso estimator.

Published in at http://dx.doi.org/10.3150/11-BEJ364 the Bernoulli (http://isi.cbs.nl/bernoulli/) by the International Statistical Institute/Bernoulli Society (http://isi.cbs.nl/BS/bshome.htm)

References in corpus (9)

• High-dimensional graphs and variable selection with the Lasso
• A note on the group lasso and a sparse group lasso
• High-dimensional Ising model selection using ${\ell }_1$-regularized logistic regression
• High-dimensional generalized linear models and the lasso
• A Unified Framework for High-Dimensional Analysis of M-Estimators with Decomposable Regularizers
• Sharp thresholds for high-dimensional and noisy recovery of sparsity
• Structured variable selection and estimation
• Best subset selection, persistence in high-dimensional statistical learning and optimization under $l_1$ constraint
• On non-asymptotic bounds for estimation in generalized linear models with highly correlated design

Cited by in corpus (5)

• A lasso for hierarchical interactions
• Maximum likelihood estimation in log-linear models
• Model selection and local geometry
• Composite mixture of log-linear models for categorical data
• Autoregressive Process Modeling via the Lasso Procedure