This section contains small machine learning exercises. Ideally each note is set to be independent and can be reused as a reference.
Logistic Regression
Binary logistic regression as a Bernoulli model: setup, cross-entropy objective, and gradient-based fitting.