01·Logistic regression — the foundation of classification

Introduction

Logistic regression is a linear probabilistic classifier: a linear combination of features is passed through the sigmoid function (or softmax for the multiclass variant), yielding a class probability. This lesson dissects log-odds (logit), the cross-entropy loss, gradient-based optimization, L1/L2 regularization, and the interpretation of coefficients as effects on log-odds. Despite the name "regression", it is a classifier — and still one of the most widely used baselines in industry (Hosmer & Lemeshow 2013, scikit-learn LogisticRegression).