THE METHODS GUIDE
What’s inside the models?
Ten common machine learning methods, explained simply: the idea, how each one learns, how to read what it tells you, and when to use it.
The fitted-model charts use synthetic examples from the fictional community kitchen and are calculated in your browser. Naive Bayes is illustrated with fixed fictional message counts, and neural networks with an architecture diagram. Most methods answer one of three questions: predict a number, predict a category or find groups.
Sources
These explanations and charts are original Code With Impact material. We checked them against An Introduction to Statistical Learning (James, Witten, Hastie, Tibshirani and Taylor), the scikit-learn user guide and Google’s Machine Learning Crash Course. No text, figures or code are copied from them. The data is synthetic.