Introduction to Machine Learning for Coders is a free twelve-lesson course from fast.ai, taught by Jeremy Howard, that teaches 📝Machine Learning (ML) to working developers from a code-first perspective.
The material runs about 24 hours of video and was recorded at the University of San Francisco for its Master of Science in Data Science program before being published openly by fast.ai; it is hosted in the group's 2018 course archive. Entry requires roughly one year of coding experience and high school mathematics, or the willingness to pick that math up along the way. The suggested pace is eight hours a week for twelve weeks.
The sequence starts with random forests and stays there long enough to cover training, validation, model interpretation, and feature importance before any deeper theory arrives. It then moves through gradient descent, logistic regression, regularization, and learning rates into natural language processing and embeddings, and closes on the ethical implications of deployed models. That ordering — build a working model, then explain why it works — is fast.ai's signature inversion of the conventional math-first curriculum, and it sits alongside the group's deep learning sequence as its treatment of classical machine learning.
"fast.ai... can actually get smart, motivated students to the point of being able to create industrial-grade ML deployments" ~ Harvard Business Review