Feature importance - Bayesian model with horseshoe prior

Hello everyone! I trained a Bayesian glm model with a horseshoe prior distribution. My dependent variable is categorical with two possible values (positive and negative). I am interested to know how to perform a feature importance analysis and find the features with the highest discriminating power between these two groups.

Thank you!!

have you looked at the examples at Model assesment, selection and inference after selection | avehtari.github.io ? For example, diabetes case study: Bayesian Logistic Regression with rstanarm

1 Like