MPBART - Multinomial Probit Bayesian Additive Regression Trees

TL;DR

This paper introduces MPBART, a flexible Bayesian model for multinomial probit problems that improves predictive accuracy in discrete choice and multiclass classification tasks, supported by simulations and real data examples.

Contribution

It extends Bayesian Additive Regression Trees to multinomial probit models, enabling inclusion of diverse predictors and demonstrating superior predictive performance.

Findings

MPBART outperforms existing methods in predictive accuracy

Demonstrated effectiveness through simulations and real data

Provides an accessible R package for implementation

Abstract

This article proposes Multinomial Probit Bayesian Additive Regression Trees (MPBART) as a multinomial probit extension of BART - Bayesian Additive Regression Trees (Chipman et al (2010)). MPBART is flexible to allow inclusion of predictors that describe the observed units as well as the available choice alternatives. Through two simulation studies and four real data examples, we show that MPBART exhibits very good predictive performance in comparison to other discrete choice and multiclass classification methods. To implement MPBART, we have developed an R package mpbart available freely from CRAN repositories.