Two 3-hour classes

Math concepts to teach before students trust linear regression.

This mini-site is now divided into two instructor slide packs: Session 1 covers simple linear regression and the math of fitting a line; Session 2 covers evaluation, multiple regression, and practical model risks.

Simple regression OLS Gradient descent R2 and RMSE Multicollinearity
input feature x target y

Classroom hook: every dot is reality, the line is our simplified explanation of reality.

1

Session 1

Simple linear regression: ML motivation, applications, line equation, residuals, cost functions, OLS, and gradient descent.

 2

Session 2

Multiple linear regression: evaluation metrics, business workflow, coefficients, assumptions, multicollinearity, scaling, and regularization teaser.

Suggested order for two classes

Class Math focus Student outcome
Class 1 Line equation, slope/intercept, residuals, RSS, MSE, OLS intuition, gradient descent intuition. They can explain what the line means and why one line is better than another.
Class 2 Metrics, R2, multiple regression equation, coefficient interpretation, assumptions, multicollinearity, polynomial idea. They can build and judge a regression model like a practical ML workflow.