This tutorial discusses neural networks models for general classification and regression problems. We start with the basics of feedforward networks and backpropagation algorithm. We will then introduce some of the recent developments in learning deep neural networks such as stacked autoencoders and convolutional networks. We will discuss convolutional networks for image recognition in some detail. We would also briefly discuss some recurrent network models such as Restricted Boltzman Machines. The tutorial would also contain an introduction to Python and implementation of some of the deep learning algorithms in Python on GPU-based systems. The tutorial would largely be self-contained and we would not assume any background in neural networks.