Statistics for High-Dimensional Data

2019-2020

Course Objective

Teaching students the adjustments to classical statistical methodology,
necessary to analyze high-dimensional data.

Course Content

This course gives an overview of modern statistical methods that are
used in the analysis of big or high-dimensional data. Such data usually
comprise a limited number of individuals that have been characterized
with respect to many traits. These data arises in genomics, where
genetic information is measured for many thousands of genes
simultaneously, in functional MRI imaging of
the brain, but also in economic applications.

The course covers some of the most important statistical issues for big
or high-dimensional data, including: a) multiple testing, the
family-wise error rate and false discovery rate control; b) shrinkage,
Stein's estimator; c) penalized estimation, in particular ridge and
lasso regression; and (time-allowing) d) penalized estimation of
covariance matrices.

Several types of high-dimensional data will be discussed and used as
examples during the course. In terms of applications the course focuses
on cancer genomics, but theoretical aspects will apply to other fields
as well.

Teaching Methods

Lectures + exercises

Method of Assessment

Written exam

Literature

"Tutorial in biostatistics: multiple hypothesis testing in genomics" by
Goeman & Solari (article in Statistics in Medicine)

"Lecture notes on ridge regression" by van Wieringen (available via
Arxiv)

Plus additional handouts provided by the lecturer.

Target Audience

mMath, mSFM

Custom Course Registration

This course is offered every other year. The next opportunity to enroll will be 2019-2020.

Recommended background knowledge

The course requires familiarity with statistics, generalized linear
models, linear algebra, probability and basic analysis.

General Information

Course Code X_405113
Credits 6 EC
Period P4+5
Course Level 400
Language of Tuition English
Faculty Faculty of Science
Course Coordinator dr. W.N. van Wieringen
Examiner dr. W.N. van Wieringen
Teaching Staff dr. W.N. van Wieringen

Practical Information

You need to register for this course yourself

Teaching Methods Lecture
Target audiences

This course is also available as: