Table of Contents

Winter 2020

Syllabus.

(see also the course webpage)

Regression and Data Analysis

STATS 513-WIN 2020
MF 8:30-10:00am 296WEISER

Instructor: Xianshi (Sean) Yu

​ Email: xsyu@umich.edu

​ Office hours: Tue 4:00-5:30pm, Conference on Canvas

GSI: Rebeka Man

​ Email: mrebeka@umich.edu

​ Office hours: Thu 12:30-2:00pm, Conference on Canvas

Course Outline:

  • Linear models (basic): definition, fitting, interpretation of results
  • Linear models (advanced): variable selection, transformations, diagnostics, outliers and influential observations, model selection, shrinkage methods
  • a complete example with real data

logistic regression is selective, depending on the progress of the course

Text Books:

Julian Faraway (2014) Linear Models with R, 2nd edition. Julian Faraway (2016) Extending the Linear Models with R, 2nd edition. (Supplementary)

Computing:

The software R will be used for this course, which is free for both Windows and Mac. Instructions about the basic implementations of R will be included in the lectures.

Prerequisite: Knowledge of matrix algebra, introductory probability, and mathematical statistics.

Grading: Homework (30%), Midterm (30%), Final (40%)

Homework: R is the required software for completing the homework. Late homework is not accepted. If you are unable to attend class on the day a homework is due, e-mail it to the grader (the GSI) before the due date. The worst homework score will be dropped in calculating the final score.

Exams:

  • Midterm: 03/16 Mon 8:30-9:50am, in class.
  • Final: 04/23 Thu 1:30-3:30pm (not cumulative)

Arrangements for virtual instruction mode (in effect from 03/12/2020)

  • Lectures are given virtually through BlueJeans. This is the link to the lecture. You can also click BlueJeans on Canvas to join the lecture.
  • Lecture videos will be put on Media Gallery. Canvas will take some time to process the video before you can see it there.
  • Office Hours are held on Conference on Canvas. You can join the conference by clicking Conference and then click join.
  • Homework is to be submitted online through Canvas. Please click Assignments, select the corresponding assignment and then click Submit Assignmentt to upload your electronic file. You could also choose to email your homework to the GSI if you prefer this.