Analyzing Restricted Boltzmann Machine Neural Network for Building Recommender Systems

Deep learning becomes a useful technology in many fields for solving many problems with high ratio of user satisfaction as well playing a big role as alternative smart techniques for handling big data problems. While recommender system one of the hottest fields that deep learning achieves huge success in case of predication the user needs based on their preferences. Restricted Boltzmann machine (RBM neural network) used for building recommender system with various types of the recommender. This paper analyzes the factors (Batch size, and number of epochs) that affect RBM to build a such recommender while shows the errors and time consumption as analysis factors on a dataset with 1 million movie rating.