Logistic Regression Review
Nov 21, 2024
Announcements
Lab 06 and HW 04 due TODAY at 11:59pm
Statistics experience due Tuesday, November 26
Team Feedback #2 (from Teammates) due Tuesday, November 26
Project:
Project meetings: November 25 and 26
- Click here to sign up (1 slot per team) by November 22
Project: Draft report due + peer review December 2
Round 1 submission (optional) due December 8 at 11:59pm
Lecture recordings available until December 5 at 11:45am (link on sidebar of course website)
Exam 02 format
50 points total
in-class: 38 points
take-home: 12 points
In-class: 75 minutes during Thursday, December 5 lecture
Take-home: due Friday, December 6 at 11:59pm (grace period: can submit without late penalty until Saturday, December 7 at 11:59pm)
Need a note from your academic dean if you miss any part of the exam
Exam 02 content
Concepts from the first half of the semester continue to apply, but the exam will focus on new content since Exam 01.
Multiple linear regression
Maximum likelihood estimation
Model diagnostics
Multicollinearity
Variable transformations
Model comparison
Logistic regression
Probabilities, odds, odds ratios
Maximum likelihood estimation
Predicted probabilities and classes
ROC curve and AUC
Inference
Assumptions
Not on the exam: Newton-Raphson method
Tips for studying
Rework derivations from assignments and lecture notes
Review exercises in AEs and assignments, asking “why” as you review your process and reasoning
- e.g., Why do we include “holding all else constant” in interpretations?
Understand similarities and differences between linear and logistic regression
- How are interpretations for logistic regression similar to interpretations for linear regression with response \(\log(y)\)? How are they different?
Focus on understanding not memorization
Explain concepts / process to others
Ask questions in office hours
Review lecture recordings as needed
