Linear & Logistic Regression

Learning Outcomes

Understand how to fit and interpret linear regression models.
Understand logistic regression and probability-based classification.
Evaluate model performance using R², MSE, accuracy, precision, recall, F1, and ROC curves.
Understand regularization and the bias-variance tradeoff.
Connect regression and classification to optimization and gradient descent.
Apply regression and classification techniques to real datasets using Python.

Statistical Learning with Python (ISLP) , Chapter 2.1-2.2: Statistical Learning and Model Accuracy
Statistical Learning with Python (ISLP) , Chapter 3.1-3.5: Linear Regression
Statistical Learning with Python (ISLP) , Chapter 4.1-4.3: Logistic Regression
ISLP Chapter 6: Linear Model Selection and Regularization

Date	Key Topics	Meeting Notes and Lesson Objectives	Assignments
10 Feb 2026	Lesson 05: Linear Regression	Meeting notes linked here. Learn to: (1) apply linear regression, (2) use cross-validation, (3) understand geometric, statistical, and optimization perspectives on least squares, and (4) derive the normal equations.	ISLP 69-108; MLP Mod 2.
17 Feb 2026	Lesson 06: Logistic Regression & Classification	Meeting notes linked here. Understand logistic regression, probabilistic classification, confusion matrices, precision, recall, F1, ROC curves, and the relationship between training loss and evaluation metrics.	ISLP 135-167; MLP Mod 3.
24 Feb 2026	Lesson 07: Model Selection & Regularization	Meeting notes linked here. Understand shrinkage, Ridge regression, Lasso, model-selection criteria, and the role of regularization in controlling variance.	ISLP 229-253; GitHub Introduction.
03 Mar 2026	Lesson 08: Classification Metrics & Gradient Descent	Meeting notes linked here. Understand classification metrics and gradient descent. Connect the model, loss function, gradient, learning rate, and optimization algorithm.	Géron 89-111.
10 Mar 2026	Lesson 09: Classification Model Comparison	Meeting notes linked here. Compare logistic regression, LDA, SVM, and KNN. Understand how model parameters change bias, variance, flexibility, computational cost, and interpretability.	ISLP 173-188.