Analysis of deep learning models using convolution neural network techniques

Deep Learning is the one of the souls of Artificial Intelligence and it is rapid growing in the medical data analysis research field, in many conditions Deep learning models are look like the neurons in brain, although both contain enormous number of computation Neurons units also called neurons that are not extremely intelligent in separation but improve optimistically when they interact with each other. The key objective is that many Convolution Neural Network models are available for image analysis which gives different accuracy in different aspects by training the model. A major analysis of Convolution models using Multilayer Perceptron is driven to analyses the image dataset of handwritten digits and to experiment by variations that are occurred in during the various changes that applied to the Convolution techniques like padding, stride and pooling to get best models in terms of the best accuracy and time optimization by minimizing the loss function.