The horseshoe estimator for sparse signals

作者：C. M. Carvalho, Nick Polson, James G. Scott · 发表于：Biometrika · 年份：2010 · DOI：10.1093/biomet/asq017 · 被引用次数：1456 · 研究领域：Statistical Methods and Inference、Advanced Statistical Methods and Models、Statistical Distribution Estimation and Applications

This paper proposes a new approach to sparsity, called the horseshoe estimator, which arises from a prior based on multivariate-normal scale mixtures. We describe the estimator’s advantages over existing approaches, including its robustness, adaptivity to different sparsity patterns and analytical tractability. We prove two theorems: one that characterizes the horseshoe estimator’s tail robustness and the other that demonstrates a super-efficient rate of convergence to the correct estimate of the sampling density in sparse situations. Finally, using both real and simulated data, we show that the horseshoe estimator corresponds quite closely to the answers obtained by Bayesian model averaging under a point-mass mixture prior.