This course starts with the fundamentals of statistics and logic before progressing to discussing more applied, specific uses of AI, including robotics, computer vision, and natural language processing. It is taught by two experienced AI researchers, Peter Norvig and Sebastian Thrun, and is designed to take around four months to complete.
Machine Learning – Stanford University (Coursera)
Often cited by AI experts as the single most important online resource for anyone wanting to learn AI, this course is led by Andrew Ng, who founded Google’s pioneering Google Brain deep learning program. This course gives a thorough grounding in the mathematical, statistical, and computer science fundamentals that go into developing and deploying automated learning machines.
AI for Everyone – Andrew Ng (Coursera)
Another course from Andrew Ng – this one explicitly aimed at those who don’t need an in-depth technical understanding of the subject but who may want to begin leveraging AI in their organizations or working to roll out AI initiatives while working with non-technical teams. It covers the workflow of running AI projects as well as how to develop a strategy around AI deployments in business.
This course is also listed in my guide to the best free data science courses. Of course, there’s a great deal of crossover between the two subjects, as data science is the foundation of all of today’s AI. If you’re confused about the terminology, then think of machine learning as a technique that leverages data science to work towards achieving what we currently understand as AI. This course gives a great overview as it starts by explaining the core data science concepts before moving on to demonstrate how they are applied in machine learning.
Machine Learning Crash Course – Google
Another Google course and this one is said to be required reading for everyone whose work is involved with AI at the tech giant. This course covers the basics but also moves onto the theory and practical applications of TensorFlow, Google’s open-source deep-learning library that it uses in many of its own AI-enhanced services and projects.
Learning From Data (Introductory Machine Learning) – Caltech (EdX)
Starting with theoretical principles such as “what is learning?” and “can machines learn?” this course covers advanced practical applications including creating ML algorithms used to power neural networks. It aims to help those who are set on a career as a data scientist or analyst. Like many of the courses covered here, all of the materials are freely available, but you can pay $50 for official certification at the end.
Another course that takes a slightly different approach, here you are taken through the practical steps necessary to build machines that solve a number of real-world AI problems, such as driving a car or playing a game. It also covers Q-learning, a form of machine learning based on reinforcement learning, that is gaining in popularity in cutting-edge applications.
Creative Applications of Deep Learning With Tensorflow – Kadenze (Class Central)
Deep learning is one of the most advanced fields of AI, and one that is pushing the boundaries of creating machines that can think and learn like humans. This is another course focused on the open-source TensorFlow framework originally created by Google for use in Deep Learning and is one that has received good reviews for giving an easy-to-follow guide to a complex technical subject.