Courses

Online Content: lecture notes, homework assignments, programming assignments

This course gives an in depth overview of both the foundational ideas and the recent advances in neural net algorithms. Additionally, the course using PyTorch throughout, provided incredibly useful practical knowledge

Content provided: lecture notes

Unsatisfied with the lack of mathematical rigor and wanting a more up to date understanding of machine learning, I took Stanford’s CS229. A lot of the topics covered are the same as Coursera’s Machine Learning course, but Stanford’s course goes into a lot more detail.

Content provided: lecture notes and problem sets

Probability is a vital concept for anyone pursuing CS, let alone AI. I was never able to study probability in highschool but I knew I needed to learn it. Oxford’s first year probability course is develops the concept of chance in a mathematical framework. I highly recommend checking it out if you feel like you need to touch up on probability.

The Google Machine Learning Crash Course is a quick course that is useful for beginners. It provides the intuition behind the concepts and has coding assignments that teach you how to use the TensorFlow library.

Coursera’s Machine Learning provides a great introduction to the industry application machine learning concepts and is taught by an incredible instructor, Andrew Ng. The course is slightly outdated and I recommend not paying for the programming assignments as they use MatLab, which is barely used anymore for machine learning.

Content provided: lecture notes, GitHub repository, assignment handouts

CS231n provides a great overview on the intuition behind convolutional neural networks and how to implement them in python.