Module 3: ML Classifiers with Scikit-Learn

A guided tour of the workhorse classifiers in scikit-learn — how each one draws its decision boundary, what its hyperparameters control, and when to reach for it in practice.

What You’ll Learn¶

• Logistic regression and probabilistic classification

• Support vector machines, maximum margin, and kernels for nonlinear problems

• Decision trees and how impurity measures drive splits

• K-nearest neighbors as a lazy, instance-based learner

• Regularization and the bias–variance tradeoff across model families

Lecture Materials¶

• 📝 Lecture Notes (PDF)

• 🧮 Math Notes (PDF)

• 📓 Chapter notebook: A Tour of Machine Learning Classifiers Using Scikit-Learn

Lecture Videos¶

Part 1: Introduction and Choosing a Classification Algorithm | CAP 6606 — Module 3, Part 1¶

Part 2: Training a Perceptron with Scikit-Learn | CAP 6606 — Module 3, Part 2¶

Part 3: Logistic Regression | CAP 6606 — Module 3, Part 3¶

Part 4: From Adaline to Logistic Regression: Training with Scikit-Learn | CAP 6606 — Module 3, Part 4¶

Part 5: Key Concepts and Regularization | CAP 6606 — Module 3, Part 5¶

Part 6: Support Vector Machines and Kernel Methods | CAP 6606 — Module 3, Part 6¶

Part 7: Decision Trees and Random Forests | CAP 6606 — Module 3, Part 7¶

Part 8: K-Nearest Neighbors and Module Summary | CAP 6606 — Module 3, Part 8¶

Prerequisites¶

• Module 2 (gradient descent, decision boundaries)

Reference¶

Raschka & Mirjalili, Machine Learning with PyTorch and Scikit-Learn, Chapter 3.