Methods Week in Statistics: Robust Statistics and Outlier Detection

The Center for Statistics in Oldenburg and Bremen (ZeSOB) organized its second

Methods Week in Statistics on the topic: “Robust Statistics and Outlier Detection”

from September 20th to 23th, 2017, at the University of Oldenburg, Building W1, Lecture Hall 017,

The course, offered at own costs, was targeted to practitioners and application-oriented statisticians. We presented a variety of procedures for outlier detection, robust alternatives to classical procedures, and corresponding diagnostics, together with their software implementation and their methodological background. Focus was on applications and on the appropriate choice of methods for concrete model classes and problems.

 

Lecturers:
Peter Ruckdeschel, Carl von Ossietzky University of Oldenburg
Valentin Todorov, United Nations Industrial Development Organization

Course members

Program of the ZeSOB Methods Week in Statistic on “Robust Statistics and Outlier Detection”:

Wednesday, Sep 20th 2017

13.45h Welcome (ZeSOB team)

14.00h – 17.00h (with 30 minutes coffee break) Introduction to Robust Statistics (PR)

Thursday, Sep 21st 2017

09.00h – 12:30h (with 30 minutes coffee break) Hands-on with data examples and R packages on Robust Statistics (PR&VT)

12.30h – 14.00h Lunch break 14.00h – 17.00h (with 30 minutes coffee break) Robust Statistics for Multivariate Data and Regression (VT)

18.00h – 22.00h Dinner in “Drögen Hasen” (approx. 15 minutes walk)

Friday, Sep 22 nd 2017

09.00h – 12.30h (with 30 minutes coffee break) Hands-on R data examples, diagnostics packages on Robust Statistics (VT)

12.30h – 14.00h Lunch break 14.00h – 17.00h (with 30 minutes coffee break) Influence Curve Based Diagnostics (PR)

Saturday, Sep. 23 rd 2017

09.00h – 12.30h (with 30 minutes coffee break) Robustness in Time Series Context (PR)

12.30h – 14.00h Lunch at “Ali Baba’s” (optional)