Seminar in Statistics: Functional Data Analysis

Short Description:

Functional data consist of samples of curves or surfaces. We will study the extension of basic statistical methods to this infinite setting, and also techniques which are specific to functional data.

Prerequisites:

We require at least one course in statistics in addition to the 4th semester course Introduction to Probability and Statistics.

Program:

The schedule can be found here (UPDATED on Mar 15th).

Terms and conditions:

Each student is required to hold a presentation about a given subject. We allow 2 students per class who present 80 minutes in total (including questions). Each group will coordinate their oral contributions. Additionally, students should be present at the other seminar classes.

Two weeks prior to the presentation date each student attends a meeting with one of the lecturers or assistants to discuss the topic and discuss remaining questions. Standard time for this first meeting is Monday 13:45 at the advisor's office. It is expected that the student has already read the material and is familiar with the topic. One week prior to the presentation each student meets again with the same lecturer or assistant (they have to agree on a time). It is expected that presentation and handout are more or less finished. Both meetings will take place either on Mondays or Tuesdays.

We regard feedback as an important part of such a seminar. Rules for giving and receiving feedback can be found here, criteria for feedback can be found here. Take your talk number add 4 and consider the result modulo 12. This is when your group is supposed to give feedback.

Ask many questions! Take your talk number, substract 4 and consider the result modulo 4. This is, when you are supposed to ask questions. You do not have to prepare anything for this but each member of your group has to ask at least one question. You are encouraged to ask questions during other talks, too, of course.

Literature:

• The seminar is based mainly on the book "Functional Data Analysis" by J. O. Ramsay and B. W. Silverman, 2nd ed., Springer (2005).
• For more practical guidance, one may also have a look at "Functional Data Analysis with R and MATLAB" by J.D. Ramsay, G. Hooker and S. Graves.
• For more details on regression, regularization, PCA, kernel methods on NON-functional data see "The Elements of Statistical Learning" by T. Hastie, R. Tibshirani and J. Friedman, Springer (2009).
  

All those books are available online via the ETH library.

For specific topics, additional material can be found in the schedule under program.

Guidelines for the Presentation:

• The main goal of a talk is to impart (=vermitteln) knowledge.
• Therefore: use examples and pictures.
• There is a break of ca. 5 minutes.
• You are responsible for scheduling your presentation. Including a discussion at the end (5 minutes), you should not go over 80 minutes.
  
• After the presentation there will be feedback (5 minutes).
• Look at other sources (books, papers, internet) to gain even more knowledge about your topic.

Slides:

can be downloaded here (password required). (FINAL VERSION)

Summary by H.R. Künsch from May 26th.