English

On The Sparse Bayesian Learning Of Linear Models

Methodology 2015-02-12 v1

Abstract

This work is a re-examination of the sparse Bayesian learning (SBL) of linear regression models of Tipping (2001) in a high-dimensional setting. We propose a hard-thresholded version of the SBL estimator that achieves, for orthogonal design matrices, the non-asymptotic estimation error rate of σslogp/n\sigma\sqrt{s\log p}/\sqrt{n}, where nn is the sample size, pp the number of regressors, σ\sigma is the regression model standard deviation, and ss the number of non-zero regression coefficients. We also establish that with high-probability the estimator identifies the non-zero regression coefficients. In our simulations we found that sparse Bayesian learning regression performs better than lasso (Tibshirani (1996)) when the signal to be recovered is strong.

Keywords

Cite

@article{arxiv.1502.03416,
  title  = {On The Sparse Bayesian Learning Of Linear Models},
  author = {Yves Atchade and Chia Chye Yee},
  journal= {arXiv preprint arXiv:1502.03416},
  year   = {2015}
}

Comments

23 pages, 9 figures

R2 v1 2026-06-22T08:27:52.690Z