Lecture 18 – Logistic Regression, Part 1

by Suraj Rampure (Summer 2020)

• slides
• video playlist
• code
• code HTML

Make sure to complete the Quick Check questions in between each video. These are ungraded, but it’s in your best interest to do them.

	Video	Quick Check
18.1 Classification, and a brief overview of the machine learning taxonomy.		18.1
18.2 Pitfalls of using least squares to model probabilities. Creating a graph of averages to motivate the logistic regression model.		18.2
18.3 Deriving the logistic regression model from the assumption that the log-odds of the probability of belonging to class 1 is linear.		18.3
18.4 Formalizing the logistic regression model. Exploring properties of the logistic function. Interpreting the model coefficients.		18.4
18.5 Discussing the pitfalls of using squared loss with logistic regression.		18.5
18.6 Introducing cross-entropy loss, as a better alternative to squared loss for logistic regression.		18.6
18.7 Using maximum likelihood estimation to arrive at cross-entropy loss.		18.7
18.8 Demo of using scikit-learn to fit a logistic regression model. An overview of what's coming next.		18.8