Data augmentation for non-Gaussian regression models using variance-mean mixtures
arXiv:1103.5407
Abstract
We use the theory of normal variance-mean mixtures to derive a data-augmentation scheme for a class of common regularization problems. This generalizes existing theory on normal variance mixtures for priors in regression and classification. It also allows variants of the expectation-maximization algorithm to be brought to bear on a wider range of models than previously appreciated. We demonstrate the method on several examples, including sparse quantile regression and binary logistic regression. We also show that quasi-Newton acceleration can substantially improve the speed of the algorithm without compromising its robustness.
Added a discussion of quasi-Newton acceleration
References in corpus (5)
- Least squares after model selection in high-dimensional sparse models
- Asymptotic properties of bridge estimators in sparse high-dimensional regression models
- Cross-Fertilizing Strategies for Better EM Mountain Climbing and DA Field Exploration: A Graphical Guide Book
- Multinomial Inverse Regression for Text Analysis
- Local shrinkage rules, Levy processes, and regularized regression