(a) Two friend. Ravi and Vira, while experimenting with a deep neural network, found that their model is over-fitting the given training data. The friends referred to the available literature and materials and came up with a solution to reduce the over-fitting. Ravi decided to add drop out along with batch normalization to his model. Whereas Vira decided to go with a deeper model and employ early stopping. Assess whether the decisions Ravi and Vira are taking helps them to proceed in the right direction. Justify your answers. [3] (b) Two fricnds Mira and Rima were working on a deep neural classifier. While training, Mira's classifier obtains a training set error of 25% and validation set error of 27%; whereas Rima obtained a training set error of 4.5% and validation set error of 6%. Examine the classifiers Minh and Rirna haye designed and suggest one technique they should use to modify their dassifier if required. Justify your answers. [3]