|
Instructors: Prof. Dr.-Ing. Michael Muma
Event type:
Exercise
Org-unit: Dept. 18 - Electrical Engineering and Information Technology
Displayed in timetable as:
E:Rob SP w Biomed A.
Subject:
Crediting for:
Hours per week:
1
Language of instruction:
Englisch
Min. | Max. participants:
- | -
Course Contents:
1. Robust Signal Processing and Learning
o Measuring robustness
o Robust estimation of the mean and the variance
o Robust regression models
o Robust filtering
o Robust location and covariance estimation
o Robust clustering and classification
o Robust time-series and spectral analysis
2. Biomedical Applications
o Body-worn sensing of physiological parameters
o Electrocardiogram (ECG)
o Photoplethysmogram (PPG)
o Eye research
o Intracranial Pressure (ICP)
o Algorithms for cardiac activity monitoring
The lecture covers fundamental topics and recent developments in robust signal processing. Unlike classical signal processing, which relies strongly on the normal (Gaussian) distribution, robust methods can tolerate impulsive noise, outliers and artifacts that are frequently encountered in biomedical applications. Robust signal processing and biomedical application lectures alternate. Exercises revise the theory and apply robust signal processing algorithms to real world data.
Literature:
A manuscript and lecture slides can be downloaded via Moodle. Further reading
• Zoubir, A. M. and Koivunen, V. and Ollila, E. and Muma, M.: Robust Statistics for Signal Processing. Cambridge University Press, 2018.
• Zoubir, A. M. and Koivunen, V. and Chackchoukh J, and Muma, M. Robust Estimation in Signal Processing: A Tutorial-Style Treatment of Fundamental Concepts. IEEE Signal Proc. Mag. Vol. 29, No. 4, 2012, pp. 61-80.
• Huber, P. J. and Ronchetti, E. M.: Robust Statistics. Wiley Series in Probability and Statistics, 2009.
• Maronna, R. A. and Martin, R. D. and Yohai, V. J.: Robust Statistics: Theory and Methods. Wiley Series in Probability and Statistics, 2006.
Preconditions:
Fundamental knowledge of statistical signal processing
|
