3 years ago

On the Beta Prime Prior for Scale Parameters in High-Dimensional Bayesian Regression Models.

Malay Ghosh, Ray Bai

We study high-dimensional Bayesian linear regression with a general beta prime distribution for the scale parameter. To avoid misspecification of the hyperparameters and to enable our prior to adapt to both sparse and dense models, we propose a data-adaptive method for estimating the hyperparameters in the beta prime density. Our estimation of hyperparameters is based on maximization of marginal likelihood (MML), and we show how to incorporate our estimation procedure easily into our approximation of the posterior. Under our proposed empirical Bayes procedure, the MML estimates are never at risk of collapsing to zero. We also investigate the theoretical properties of our prior. We prove that careful selection of the hyperparameters leads to (near) minimax posterior contraction when $p \gg n$. As an aside, we also derive conditions for minimax posterior contraction under our prior for the sparse normal means problem. Finally, we demonstrate our prior's self-adaptivity and excellent finite sample performance through simulations and analysis of a gene expression data set.

Publisher URL: http://arxiv.org/abs/1807.06539

DOI: arXiv:1807.06539v2

You might also like
Discover & Discuss Important Research

Keeping up-to-date with research can feel impossible, with papers being published faster than you'll ever be able to read them. That's where Researcher comes in: we're simplifying discovery and making important discussions happen. With over 19,000 sources, including peer-reviewed journals, preprints, blogs, universities, podcasts and Live events across 10 research areas, you'll never miss what's important to you. It's like social media, but better. Oh, and we should mention - it's free.

  • Download from Google Play
  • Download from App Store
  • Download from AppInChina

Researcher displays publicly available abstracts and doesn’t host any full article content. If the content is open access, we will direct clicks from the abstracts to the publisher website and display the PDF copy on our platform. Clicks to view the full text will be directed to the publisher website, where only users with subscriptions or access through their institution are able to view the full article.