Applied Statistical Regression

Beginning of lectures: 21/09/205

Attendance certificate conditions: None. For obtaining the 5 ETCS credit points for this course, one needs to attend and pass the exam.

Doctoral students: PhD students who are after ETH credit points need to sign up with the lecturer at the beginning of the semester and hand in 5 well-solved exercises. 

Content

Here is a link to the exercises. 

Abstract

This course offers a practically oriented introduction into regression modeling methods. The basic concepts and some mathematical background are included, with the emphasis lying in learning "good practice" that can be applied in every student's own projects and daily work life.

Content

The course starts with an introduction to the regression problem and then proceeds to parameter estimation, inference, prediction and residual analysis on the basis of the simple regression model. This will be followed by a thorough discussion of variable transformations and their impact to the fitted models. We will then extend to the multiple linear regression model, for which fitting, inference and prediction will be explained. A special focus lies on diagnostic techniques, as these are at the root of fitting good models. We will then discuss the multicollinearity issue and focus on variable selection techniques, before we conclude this section with some general strategies for regression modeling. In a third part of this course, extensions of the linear model will be highlighted. This includes the flexible generalized additive model (GAM), as well as generalized linear models (GLM) on the example of binary and count response.

Objective

The students acquire advanced practical skills in linear regression analysis and are also familiar with its extensions to generalized linear modeling.

Notice

The exercises, but also the classes will be based on procedures from the freely available, open-source statistical software package R, for which an introduction will be held. See the exercise section for more information.

Documents

• Course Organization
  
• Script (version Nov 19)
• Script Chapter 4 (Dec 06)
  
• Slides Sep 21
  
• Slides Oct 12
  
• Slides Nov 30
• Slides Dec 06

Old exams

• So11 (Solution: So11) (updated Jan 21)
  
• Wi11 (Solution: Wi11) (updated Jan 21)

Literature

There are many books that cover the topics of our course. Here are some recommendations:

1. Linear Models with R, Julian J. Faraway, Chapman & Hall/CRC (2005). ISBN-10: 1584884258. 229 pages, ca. 70$.
   There is a freely available version on CRAN, entitled Practical Regression and Anova using R: http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf. This free version is not identical to the book, but it is still a very good reference. For the later chapters of the course, the second volume of Faraway’s regression literature is required:
2. Extending the Linear Model with R, Julian J. Faraway, Chapman & Hall/CRC (2006). ISBN-10: 158488424X. 312 pages, ca. 75$.
3. Applied Regression Analysis, N. Draper and H. Smith, Wiley Interscience, 3rd Edition (1998). ISBN-10: 0471170828. 736 pages, ca. 100$
4. Introduction to Linear Regression Analysis, D. Montgomery, E. Peck, G. Vining, Wiley-Interscience, 4th Edition (2006). ISBN-10: 0471754951. 640 pages, ca. 85$.
5. Applied Regression Analysis and Generalized Linear Models, J. Fox, Sage Publications, 2nd Edition (2008). ISBN-10: 0761930426. 688 pages, ca. 82$.