In this session, we will utilize Keras to address a simple image classification problem using deep learning techniques. We will begin with a multi-layer perceptron and progress to a convolutional neural network. Along the way, we will gain an understanding of stochastic gradient descent, the process of fitting neural networks, and the use of dropout for regularization of neural networks. As a practical application, we will also have the opportunity to work on a take-home problem to further develop our newfound skills through discussion on our Discord server.
Prerequisites: You need a working knowledge of Python, NumPy, Pandas and Basic ML (using Scikit-learn) to successfully complete this workshop.