This project aims to explore empirical Bayes multiple regression methods. Here is a collection of occassional unrefined Jupyter Notebooks from my daily work. For my other notebook collections, visit here.

Demonstration of EM-VAMP with changepoint simulation • Apr 9, 2021

Demonstration of EM-VAMP with adaptive shrinkage (ash) prior • Mar 31, 2021

Multiple regression with product of normals • Feb 19, 2021

Multiple regression (two predictors) with product of normals • Jan 28, 2021

Simple regression with product of normals • Jan 27, 2021

Effect of $\sigma_b$ and $\sigma_w$ in EBMR with product of normals • Jan 11, 2021

Factorization of var(W) in EBMR with product of normals • Jan 11, 2021

Check ELBO for VEB with product of two normals in case of simple regression • Jan 11, 2021

Initialization in EBMR with product of normals • Jan 7, 2021

Symmetric updates in EBMR with product of normals • Jan 6, 2021

Sequence of updates in EBMR with product of normals • Jan 5, 2021

Mr.ASH penalized regression trendfiltering demo • Jun 15, 2021

Comparison of different methods for changepoint estimation • Apr 12, 2021

Comparison of prediction accuracy of EM-VAMP, Mr.ASH and EBMR • Mar 29, 2021

Comparison of prediction accuracy of linear regression methods • Mar 24, 2021

EBMR with adaptive shrinkage prior • Dec 14, 2020

EBMR with product of coefficients • Dec 30, 2020

Peter's example of varbvs surface visualization • Sep 16, 2021

Comparison of Lasso from R (glmnet) and Python (sklearn) • Sep 20, 2021

Optmization of the penalty function using Lagrange multiplier • Jan 10, 2022

Illustration of Mr.ASH and Mr.ASHPen on some special cases • Sep 13, 2021

Demonstration of Mr.ASH penalized regression • Aug 24, 2021

Bayes Lasso using EBMR • Dec 1, 2020

Basic comparison of ridge regression methods • Nov 2, 2020