Neural networks with extensively deep architectures typically contain millions of parameters, making them both computationally expensive and time-consuming to train. In this tutorial, we'll achieve state-of-the-art image classification performance using
