Module 8: Deep Learning

Topic 1: Neural Networks

Neural networks

Before jumping into deep learning, we will learn about fully connected neural networks as they are an integral part of deep learning. We are NOT going to cover these in detail and we are not jumping into how you train them precisely. Because there is a class that covers neural networks in depth at OU and another that focuses entirely on deep learning (plus this is an overview AI class), we are going to focus on the general idea of how neural networks work for this topic.

The reading in the book is Section 21.1 (Introduction through Simple Feedforward Networks). The book jumps into a lot more detail later on, and as I said, we are not doing that for this class. But you are welcome to read through it if you want to learn more!

General neural net overview

Code.org is a really near organization that has videos online for teaching most CS topics. I really like their approach in general with their videos and thought it would be a fun one to start with. They recently put a whole series online for students to learn AI or for teachers to learn how to teach AI (they are primarily focused on K-12 plus teachers but that doesn’t mean we can’t use them as an overview!). So enjoy this quick overview video on neural nets before we dive in a bit more!

My quick overview for Neural nets

I also have a quick overview for neural nets, also a bit of an overview for the module.

Links discussed in my video

Multi-layer feedforward networks

Without jumping into the math needed to train the networks, I give you an overview of how we chain neurons together to make a neural network and to be able to approximate non-linear functions.

Copy of my slides

Exercise

Complete the exercise on neural networks