Student Seminar in Statistics

Topic: Robust Statistics

Description:

From the blurb of Robust Statistics by Maronna et al.:

«Classical statistical techniques fail to cope well with deviations from a standard distribution. Robust statistical methods take into account these deviations while estimating the parameters of parametric models, thus increasing the accuracy of the inference. [...]»

Prerequisites:

As prerequisites, we require:

• Introductory course to statistics
• One specialised course in statistics taken in a previous semester
  (e.g., Fundamentals of Mathematical Statistics, Applied Statistical
  Regression, Computational Statistics, see http://stat.ethz.ch/education)
• Basic knowledge of the software R (www.r-project.org)

Terms and conditions:

Each student is required to hold a presentation about a given subject (2 to 3 students per class and they must rotate during the presentation). Additionally, the presence in the seminar classes is mandatory. After each seminar class a homework set consisting of questions regarding the subject of the presentation(s) will be distributed among the students. The homework has to be handed-in at the beginning of the following class in written form (at least a reasonable attempt at solving the questions is expected). The homework will have to be corrected by the students that held the corresponding presentation.

Two weeks prior to the presentation date each student has to attend a meeting with the responsible assistant (see schedule) to discuss the topic and ask open questions. It is expected that the student has already read the material and is familiar with the topic. These meetings will take place either on Monday or Tuesday (time and place have to be fixed with the assistant directly).

One week prior to the presentation at the start of class each student has to hand in their presentation (PDF/PPT or copy of hand-written notes in case of black board presentation), their homework questions and their hand-out (PDF/DOC).

The corrected handouts have to be handed in at latest on Friday.

Guidelines:

We strongly recommend to use our guidelines while preparing your presentation. Also consider the following points:

• Interact with the audience by asking questions, e.g. a poll or simple TRUE/FALSE questions, and encourage questions from the audience;
• Use examples and visual aids - they will help the audience to understand your argument;
• In the schedule for your presentation plan a discussion at the end (5 - 10 minutes).

Literature:

• Robust Statistics: Theory and Methods, Maronna R., Martin D., Yohai V., Wiley, 2006. (ebook link)
• Robust Statistics (second edition), Huber P. J., Ronchetti E. M., Wiley, 2009. (ebook link) (chapter 4 only)
• Sharpening Wald-type inference in robust regression for small samples, Koller M., Stahel W. A., Computational Statistics & Data Analysis, Volume 55, Issue 8, 1 August 2011, Pages 2504-2515. (pdf link)
  

Program:

The schedule can be found here. After the first seminar class the corresponding names will be added to the schedule.

Frequently Asked Questions:

If you still have questions regarding your presentation please read our FAQ.

Slides, Synopsis, Homework, Solutions:

See here.