A non-parametric conditional factor regression model for high-dimensional input and response

TL;DR

This paper introduces a non-parametric conditional factor regression model that effectively handles high-dimensional data by using latent factors and an Indian Buffet Process prior, improving prediction accuracy.

Contribution

The novel NCFR model combines low-dimensional latent factors with an Indian Buffet Process prior for unlimited sparse dimensions, advancing high-dimensional regression techniques.

Findings

NCFR outperforms several alternatives in prediction tasks.

The model effectively reduces dimensionality while maintaining high prediction accuracy.

Experimental results demonstrate the model's robustness and scalability.

Abstract

In this paper, we propose a non-parametric conditional factor regression (NCFR)model for domains with high-dimensional input and response. NCFR enhances linear regression in two ways: a) introducing low-dimensional latent factors leading to dimensionality reduction and b) integrating an Indian Buffet Process as a prior for the latent factors to derive unlimited sparse dimensions. Experimental results comparing NCRF to several alternatives give evidence to remarkable prediction performance.