In this exercise you will implement a convolutional neural network for digit classification. The architecture of the network will be a convolution and subsampling layer followed by a densely connected output layer which will feed into the softmax regression and cross entropy objective. You will use mean pooling for the subsampling layer. You will use the back-propagation algorithm to calculate the gradient with respect to the parameters of the model. Finally you will train the parameters of the network with stochastic gradient descent and momentum