Restricted Boltzmann Machine for Classification with Hierarchical Correlated Prior

TL;DR

This paper introduces a hierarchical correlated RBM model that incorporates class relationships and orthogonal constraints to improve classification performance, demonstrating promising results on benchmark datasets.

Contribution

It proposes a novel hierarchical correlated RBM with orthogonal restrictions, capturing interclass relationships and sharing information among classes.

Findings

Outperforms baseline models on challenge datasets.

Effectively models interclass relationships.

Reduces redundancy through orthogonal constraints.

Abstract

Restricted Boltzmann machines (RBM) and its variants have become hot research topics recently, and widely applied to many classification problems, such as character recognition and document categorization. Often, classification RBM ignores the interclass relationship or prior knowledge of sharing information among classes. In this paper, we are interested in RBM with the hierarchical prior over classes. We assume parameters for nearby nodes are correlated in the hierarchical tree, and further the parameters at each node of the tree be orthogonal to those at its ancestors. We propose a hierarchical correlated RBM for classification problem, which generalizes the classification RBM with sharing information among different classes. In order to reduce the redundancy between node parameters in the hierarchy, we also introduce orthogonal restrictions to our objective function. We test our method on challenge datasets, and show promising results compared to competitive baselines.