• KARAN SHARMA
• ANITA GANPATI

Deep Learning (DL) helps in learning from experiences and understanding the real world in terms of hierarchy. Image Classification is the key application of the deep learning. Using deep learning, system can recognize objects, faces or scene like a human eye and performs much better. Image Classification can be achieved by machine learning or deep learning. Unlike machine learning, there is no need of feature selection and feature extraction in deep learning. While using machine learning, after a certain extent of training model reached a saturation point whereas this is not a case for deep learning. So, utilization of deep learning for image classification is more effective than other approaches. In deep learning, for image classification Convolutional Neural Networks (CNNs). This study is carried out to evaluate the performance of different CNN approaches for image classification over MNIST dataset The MNIST dataset is a collection of handwritten digit images, contains 10000 testing images and 60000 training images. These models are evaluated in python using Keras and TensorFlow libraries.

Deep Learning, Convolutional Neural Networks, Image Classification, MNIST, TensorFlow, Keras

"PERFORMANCE ANALYSIS OF CNN APPROACHES FOR IMAGE CLASSIFICATION OVER MNIST DATASET", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.434-439, June-2019, Available :http://www.jetir.org/papers/JETIR1906E13.pdf

"PERFORMANCE ANALYSIS OF CNN APPROACHES FOR IMAGE CLASSIFICATION OVER MNIST DATASET", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp434-439, June-2019, Available at : http://www.jetir.org/papers/JETIR1906E13.pdf