Comparison of CNN and MLP Artificial Neural Network Models for an Optical Character Recognition Test Case

Abstract. Some different topologies and parameters for a convolutional artificial neural network (CNN) are compared with a Multi-Layer Perceptron (MLP) model, using the MNIST dataset for optical character recognition. Neural networks have been used to solve a wide variety of tasks that are difficult to solve using common rule-based programming, including computer vision and speech recognition [2]. This work used the TensorFlow library to compare some CNN models and select the two with the best accuracy among them, and was also compared with the MLP model.