KF5012: Introduction to Artificial Intelligence

Part 1: Introduction to Neural Networks

Part 2: How to set up an Anaconda Virtual Environment

It is highly recommended to use Anaconda virtual environments to develop your project. This video explains why you should use Anaconda and how to set up a virtual environment

(Optional) Part 2.2: Anaconda Set up with PyCharm

I personally do all my coding in PyCharm. It is a little more difficult to get started so this video takes you through the steps to get set up if you wish to do so

Part 3: Vectors, Matrices and Tensors

Some basic understanding of mathematical structures is necessary before we explore neural networks

Part 4: Concepts of Neural Networks

Before we dive into the detail, this video will explore some of the underlying concepts that motivate us to use neural networks over traditional machine learning

Part 5.1: Overfitting

We explore why deep networks are more prone to overfit and also discuss their difficulty to generalise to new sources of data

Part 5.2: Adversarial Attacks and Interpretability

We discuss some of the other problems that deep networks face including the challenge of explainability

Part 6: Neural Networks

This video explains what a neural network is and how it functions

Part 6.2 (Optional): Worked Example

A worked example of how a neural network might determine vaccination priorities

Part 7: Deep Neural Networks

This video explores how and why deep networks often learn more expressive features, which results in a more powerful final representation

Part 8: Backpropagation

This video explores how a neural network can automatically learn its weights