Bayesian linear regression with the horseshoe hierarchy

Version 1.21.0.0 (5.03 KB) by Statovic
This file implements Bayesian linear regression with the horseshoe hierarchy.
308 Downloads
Updated Thu, 10 Aug 2017 01:21:00 +0000

View License

This file implements Bayesian linear regression with the horseshoe hierarchy using a simple Gibbs sampler.

References:
A simple sampler for Bayesian linear regression with the horseshoe hierarchy
E. Makalic and D. F. Schmidt,
arXiv:1508.03884 [stat.CO], 2015

The horseshoe estimator for sparse signals
C. M. Carvalho, N. G. Polson and J. G. Scott
Biometrika, Vol. 97, No. 2, pp. 465--480, 2010

NOTE: A new toolbox, baysreg, which implements Bayesian regularised linear and logistic regression is now available on MATLAB Central (File ID: #60823).

Cite As

Statovic (2024). Bayesian linear regression with the horseshoe hierarchy (https://www.mathworks.com/matlabcentral/fileexchange/52479-bayesian-linear-regression-with-the-horseshoe-hierarchy), MATLAB Central File Exchange. Retrieved .

MATLAB Release Compatibility
Created with R2015a
Compatible with any release
Platform Compatibility
Windows macOS Linux

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!
Version Published Release Notes
1.21.0.0

- Updated description

1.2.0.0

- Updated description to include information on the new regularised regression toolbox
- Pre-compute X'*X for the fastmvg_rue routine. This results in speed improvements which are noticeable especially for large sample sizes
- Added an arXiv reference to the paper that describes the algorithm

1.1.0.0

-Much improved sampling of the conditional distribution for betas (implemented Rue's algorithm for n>p)
-Minor speed improvement in function fastmvg
-Cleaned up the help

1.0.0.0