ABSTRACT
Deep learning is a subject that is centred on learning and developing on its own by evaluating computer algorithms, and it has arisen as a new topic in machine learning. Machine learning is used to solve easier problems, whereas deep learning is used to handle the most complex ones. The deep learning method is quite similar to how the human brain works. It is mainly utilized to handle problems that arise in the field of image classification, language translation, speech recognition, and pattern recognition. The major goal of the research reported in this chapter is to use a deep learning technique called Deep Convolutional Neural Networks (DCNN) to classify images. Here the classification models Multilayer Perceptron (MLP), CNN, and DCNN have been reviewed in detail and the algorithms are tested on two standard datasets, like MNIST and CIFAR-10. The simulation results demonstrate promising performance on the image classification models. At the end of the discussion, a deep comparative analysis has been carried out to identify the significance of the presented models. Among the presented models, the DCNN yields better accuracy of 99.83% for MNIST and 90.14% for CIFAR-10 than other methods.
