Time Series Analysis

2019-2020

Course Objective

Students will learn various techniques for the analysis of time series.
A brief sketch of the mathematical background will enable students to
select and apply proper methods for the study of signals typically found
in the movement sciences. As examples range from kinematic and (neuro-)
physiological signals students will get well-equipped to analyze and
interpret their own experimental recordings.

Course Content

Recent advances in recording techniques and increasing data storage
capacity render time series analysis a challenge. In this course various
uni-, bi-, and multivariate methods for the study of experimental data
will be outlined and critically discussed. Statistical time-domain ap-
proaches go hand in hand with Fourier analysis, Hilbert and Gabor
transforms, wavelet decomposition, et cetera. For the multivariate ex-
tension primary focus will be on principal and independent compo-nent
analysis and on investigating recordings of whole-body kinemat-ics and
electromyographic signals. All techniques will be discussed based on
current research articles and implemented by means of nu-merical
exercises (Matlab).

Teaching Methods

36 contact hours (14 seminars, 12 practicals, 10 lectures); 124 hours
self-study

A mixture of lectures, seminars, and computer practicals. At the com-
puter students will analyze typical examples of movement-related,
temporal data like kinematic or electromyographic signals. During the
seminars, research articles on the analysis of movement dynamics will be
discussed on the basis of brief summaries written by the students
(writing assignment).

Method of Assessment

60% of the grade is determined by the written exam (essay questions).
20% is determined by the quality of the written summaries and/or oral
presentations (depends on the number of participating students but will
be announced in time via BlackBlack), and 20% by the quality of solu-
tion of the computer practicals.

Entry Requirements

Basic knowledge of Matlab is mandatory.

Literature

• C. Chatfield, The analysis of time series. An introduction,
Chapman & Hall, London, 1987 (or newer edition); advised
• Several research articles that will be provided

Additional Information

Basic knowledge of Matlab is mandatory
none

Recommended background knowledge

Basic knowledge of Matlab is mandatory

General Information

Course Code B_TIMESERANA
Credits 6 EC
Period P5
Course Level 500
Language of Tuition English
Faculty Fac. of Behavioural and Movement Science
Course Coordinator prof. dr. A. Daffertshofer
Examiner prof. dr. A. Daffertshofer
Teaching Staff prof. dr. A. Daffertshofer

Practical Information

You need to register for this course yourself

Teaching Methods Seminar, Lecture, Computer lab