Suppose you have a binary classification problem, each observation is described by a 12x12 pixels image (144 features), and that you would like to build a neural network classifier. a) (10 points) How many weights would the network contain, if you built a fully-connected neural network, composed by 1 hidden layer with 50 neurons and an output layers with 2 neurons? How many parameters? b) (10 points) How many weights would the network contain, if you built convolutional neural network with 50 convolutional neurons in the first hidden layer, each with a receptive field of 5x5 pixels, and 2 fully connected neurons in the output layer?How many parameters