Cumulative Restricted Boltzmann Machines for Ordinal Matrix Data Analysis

TL;DR

This paper introduces a novel Gaussian restricted Boltzmann machine architecture tailored for modeling ordinal matrix data, demonstrating its effectiveness in capturing latent opinions and outperforming existing collaborative filtering methods on large datasets.

Contribution

The paper presents a new RBM-based model for ordinal data, including architecture, learning, and inference procedures for vector and matrix data, expanding RBM applications.

Findings

Successfully models latent opinion profiles globally

Outperforms state-of-the-art collaborative filtering methods

Applicable to recommendation systems and reviews

Abstract

Ordinal data is omnipresent in almost all multiuser-generated feedback - questionnaires, preferences etc. This paper investigates modelling of ordinal data with Gaussian restricted Boltzmann machines (RBMs). In particular, we present the model architecture, learning and inference procedures for both vector-variate and matrix-variate ordinal data. We show that our model is able to capture latent opinion profile of citizens around the world, and is competitive against state-of-art collaborative filtering techniques on large-scale public datasets. The model thus has the potential to extend application of RBMs to diverse domains such as recommendation systems, product reviews and expert assessments.