Comparative Analysis of Restricted Boltzmann Machine Models for Image Classification

Abstract

Many applications for Restricted Boltzmann Machines (RBM) have been developed for a large variety of learning problems. Recent developments have demonstrated the capacity of RBM to be powerful generative models, able to extract useful features from input data or construct deep artificial neural networks. In this work, we propose a learning algorithm to find the optimal model complexity for the RBM by improving the hidden layer. We compare the classification performance of regular RBM use RBM() function, classification RBM use stackRBM() function and Deep Belief Network (DBN) use DBN() function with different hidden layer. As a result, Stacking RBM and DBN could improve our classification performance compare to regular RBM.