Nonparametric Bayes Classification via Learning of Affine Subspaces

TL;DR

This paper introduces a flexible non-parametric Bayesian classifier that learns a low-dimensional subspace for high-dimensional data, improving interpretability and robustness in classification tasks.

Contribution

It proposes a novel model that learns affine subspaces for dimension reduction and classification, with interpretable parameters and consistent posterior estimation.

Findings

Model effectively identifies important predictors.

Posterior estimates are shown to be consistent.

Real data applications demonstrate practical utility.

Abstract

The goal of this presentation is to build an efficient non-parametric Bayes classifier in the presence of large numbers of predictors. When analyzing such data, parametric models are often too inflexible while non-parametric procedures tend to be non-robust because of insufficient data on these high dimensional spaces. When dealing with these types of data, it is often the case that most of the variability tends to lie along a few directions, or more generally along a much smaller dimensional subspace of the feature space. Hence a class of regression models is proposed that flexibly learn about this subspace while simultaneously performing dimension reduction in classification. This methodology, allows the cell probabilities to vary non-parametrically based on a few coordinates expressed as linear combinations of the predictors. Also, as opposed to many black-box methods for dimensionality reduction, the proposed model is appealing in having clearly interpretable and identifiable parameters which provide insight into which predictors are important in determining accurate classification boundaries. Gibbs sampling methods are developed for posterior computations. The estimated cell probabilities are theoretically shown to be consistent, and real data applications are included to support the findings.