Linear and logistic regression

Linear regression predicts numbers with a weighted sum. Logistic regression predicts class probabilities with a sigmoid on a weighted sum—still a workhorse classifier.

Best first model for many problems: fast, auditable coefficients, strong baseline.

Fit weights to reduce error (squared loss or log loss) with regularization. Check residuals and calibration.

Predict a simple numeric target from one feature on a spreadsheet; then try a binary label with logistic intuition.

⚠️ Myth: Linear models can’t be useful in the deep learning era.

✓ Reality: They win on small data, speed, and interpretability needs.

• Course 03 supervised; Course 04 math
• Google ML Crash Course: https://developers.google.com/machine-learning/crash-course ↗