Welcome Aboard! 🙌

MATH 4780 / MSSC 5780 Regression Analysis

Dr. Cheng-Han Yu
Department of Mathematical and Statistical Sciences
Marquette University

Taipei, Taiwan

• Hello everyone, how are you? I hope you have a great winter break. Welcome to 3570 Introduction to Data Science course. I am your instructor Cheng-Han Yu.

• First thing first. Get to know each other. Let me first introduce my self. I was born and grew up in Taipei, Taiwan, my home country, which is a island right next to China. Taipei is a big city in terms of population. It is about the same population size as Chicago. This building is called Taipei 101, which is the tallest building in Taiwan.

My Journey

• Assistant Professor (2020/08 - )

• Postdoctoral Fellow

• PhD in Statistics

• MA in Economics/PhD program in Statistics

After college, working and doing military service for several years, I came to the US for my PhD degree. Originally I would like to study economics, but then I switched my major to statistics.

• I got my master degree in economics from Indiana University Bloomington, then I transferred to UC Santa Cruz to finish my PhD studies.
• Then I spent two years doing my postdoctoral research at Rice University in Houston, Texas.
• Finally, in fall 2020, I came to Marquette as an assistant professor.
• Midwest/Indiana-West/California-South/Texas-Midwest/Wisconsin
• Been to any one of these universities/cities?
• The most beautiful campus.
• Who are international students? I can totally understand how hard studying and living in another country. Feel free to share your stories or difficulties, and I am more than happy to help you if you have any questions.
• Poor listening and speaking skills. I was shy.
• OK so, this is my background. How about you introducing yourself as well. You can share anything, your major, hobbies, your favorite food, what do you want to do after graduation, anything,
• I have the class list. I’d like to learn your face and remember your name. You know, you all wear a mask. It’s hard to recognize you and connect your name and your face.
• When I call your name, you can say something about yourself. No need to be long, couple of seconds are good.

My Research

• Bayesian spatio-temporal modeling and machine learning algorithms in neuroimaging
• Bayesian Deep Learning for image classification
• Efficient MCMC for big \(n\) big \(p\) streaming data
• Game-based learning for STEM and data science education

fMRI

EEG

How to Reach Me

• Office hours TuTh 3:20 - 4:50 PM in Cudahy Hall 353.
• 📧 cheng-han.yu@marquette.edu
  • Answer questions within 24 hours.
  • Expect a reply on Monday if shoot me a message on weekends.
  • Start your subject line with [math4780] or [mssc5780] followed by a clear description of your question.

• I will NOT reply your e-mail if … Check the email policy in the syllabus!

Textbook (LRA)

• Introduction to Linear Regression Analysis, 6th edition, by D. C. Montgomery, E. A. Peck and G. G. Vining. Publisher: Wiley.
• Ch 1 ~ 10, 13.
• Careful about typos.

In the Preface,

The book … for a course taken by seniors and 1st-year graduate students….

Some knowledge of matrix algebra is also necessary.

• Course materials are grabbed from several books and resources.

• Our course website would be helpful.

Reference (CMR)

• Classical and Modern Regression with Applications, by Raymond Myers. Publisher: Duxbury Press.
• Ch 1 ~ 8.
• The textbook when I was a master student at Indiana University.
• Outdated (published in 1990) and no code involved.
• Explains concepts well.

In the Preface,

The book … for seniors and graduate students majoring in statistics or user subject-matter fields.

• I also provide some other reference books. They are all optional, but some course materials may borrow from these books.
• It’s a pretty old book which was published in 1990. So it’s a little bit outdated. But it explains basic ideas and concepts pretty well. I enjoyed reading it.

Reference (CAR)

• An R Companion to Applied Regression, by John Fox and Sanford Weisberg. Publisher: SAGE.
• Ch 3, 5, 8.
• R scripts and data on the book Website
• Doing regression easier by the car and effects R packages.

• If you wanna learn how to use R to do regression, this book could be helpful.
• The authors create 2 R packages car and effects that make doing regression so much easier.
• Later you’ll see we can just use some simple function calls to generate some important and useful graphics for regression analysis.

John Fox received a BA from the City College of New York and a PhD from the University of Michigan, both in Sociology. Sanford Weisberg is Professor Emeritus of statistics at the University of Minnesota. He has also served as the director of the University′s Statistical Consulting Service, and has worked with hundreds of social scientists and others on the statistical aspects of their research. He earned a BA in statistics from the University of California, Berkeley, and a Ph.D., also in statistics, from Harvard University

Reference (ISL)

• An Introduction to Statistical Learning, by James et. al. Publisher: Springer.
• Ch 3.
• R and Python code

In the Preface,

ISL … for advanced undergraduates or master’s students in Statistics or related quantitative fields……

More References

• Applied Regression Analysis and Generalized Linear Models

• Applied Linear Regression

• Linear Models with R/Python

• Applied Linear Statistical Models

• Regression Daignostics

and more!

Prerequisites

• On bulletin: MATH 2780 (Baby Regression), MATH 4720 (Intro Stats) or equivalent.

Helpful if you

• code (any language) 💻

• learned basic linear algebra (MATH 3100) 🔢

• took probability (MATH 4700) and statistical inference (MATH 4710) 🎲

• On the Marquette bulletin, the prerequisite is either MATH 2780 or MATH 4720 or equivalent.
• But this course requires more than basic statistics experience.
• Nowadays, nobody is doing regression by hand, and we will write some code to do regression. So it would be helpful if you code any language before.

My Statistics Book

• Please review probability distributions, confidence interval, and hypothesis testing!

Course Website - https://math4780-f23.github.io/website/

• All course materials

Course Website - D2L

• Homework submission
• Gradebook

Grading Policy ✨

• For MATH 4780 (MSSC 5780) students, the grade is earned out of 1000 (1200) total points distributed as follows:
  • Homework 1 to 6: 480 pts (80 pts each)
  • In-class Exam: 160 pts
  • Final project: 300 pts
  • Class Participation: 60 pts
  • MSSC 5780 work: 200 pts
• ❌ No extra credit projects/homework/exam to compensate for a poor grade.
• Individual grade will NOT be curved.
• Want to obtain a good grade? Study hard. No pain, no gain! ✍ ✍

• Everyone has the same opportunity to do well in this class.
• 4780 students are encouraged to do MSSC problems: Up to 100 pts extra credit.
• I am not a harsh grader, but this is not an easy A course as well.

Grade-Percentage Conversion

• \([x, y)\) means greater than or equal to \(x\) and less than \(y\).

• For example, 94.0 is in [94, 100] and the grade is A and 93.8 is in [90, 94) and the grade is A-.

Grade	Percentage
A	[94, 100]
A-	[90, 94)
B+	[87, 90)
B	[84, 87)
B-	[80, 84)
C+	[77, 80)
C	[74, 77)
C-	[70, 74)
D+	[65, 70)
D	[60, 65)
F	[0, 60)

Homework (480 pts)

• Homework will be assigned through the course website in weekly schedule.

• D2L > Assessments > Dropbox and upload your homework in PDF format.

• You must submit YOUR OWN work. 🙏

• ❌ No make-up homework.

• Due Friday 11:59 PM (09/08, 09/22, 10/06, 10/27, 11/17, 12/08. Hard deadline and no late submission).

• ❌ Handwriting is NOT allowed for data analysis part.

In-class Exam (160 pts)

• Midterm exam is held in class on 10/17.

• 📄 One piece of letter size cheat sheet is allowed. It has to be turned-in with your in-class exam. (Or entirely open book?)

• Exam covers materials in Week 1 to 7.

Project (300 pts)

• You will be doing a team project.

• What is your project about?

• Written report or oral presentation or both?

• How is your projected evaluated?

• More information will be released later.

analyzing real-world data using regression methods

Class Participation (60 pts)

• Five minute presentation on exercise problems?

• More details will be released later.

MSSC 5780 Work (200 pts)

• Some extra work are for MSSC 5780 students.
• 👍 MATH 4780 students are encouraged to do the MSSC problems to earn extra points!

What Computing Language We Use (I Teach)?

• 🖖 The best language for statistical computing!

#| fig-cap: ‘R Website’ - R is the best language for statistical computing. - Its code for doing regression analysis is pretty clean and concise. - You won’t regret learning R.

Which Programming Language?

• ✅ May use any other language (Python, MATLAB, etc) to do your work.

• ❌ I would NOT debug your code or comment on your technical issues of those languages.

• If time permitted, I’ll translate R to Python.

Academic Integrity

This course expects all students to follow University and College statements on academic integrity.

• Honor Pledge and Honor Code: I recognize the importance of personal integrity in all aspects of life and work. I commit myself to truthfulness, honor, and responsibility, by which I earn the respect of others. I support the development of good character, and commit myself to uphold the highest standards of academic integrity as an important aspect of personal integrity. My commitment obliges me to conduct myself according to the Marquette University Honor Code.

• ❌ You know what I am talking about. Yes, DO NOT CHEAT.