A Comprehensive Analysis of Convolutional Neural Network Models

Abstract

Abstract. Deep learning is an emerging field of machine learning that has been grown rapidly and applies to many domains with high success frequency including image processing, speech recognition and text processing. Experiments shows its high applicability and significant performance compared with traditional machine learning approaches. Deep learning algorithms are mainly an inheritor of artificial neural network architecture with higher number of hidden layers, therefore known as deep neural networks. A type of feed forward network, Convolutional Neural Network (CNN), is one of the most widespread deep neural network models. This paper presents a brief survey of well-known and established CNN models. It also contains comparative evaluation of these models by considering different parameters.