Change Detection and Decision Theory (ASI)

Informations générales

  • ECTS

    ECTS 1

Goal(s)

- To be able to develop a real-time change detection algorithm.
 - To be able to correctly choose, tune and apply classical real-time change detection algorithms (Cusum, GLR and Brandt's GLR algorithms).

Content(s)

 1. Definitions
  - signals, abrupt changes, and real-time approach
2. Theoretical background
 - hypothesis testing and decision theory
 - estimation theory
3. Signal with independent samples
 - mean and variance change detection
 - CUSUM, GLR and Brandt's GLR algorithms
 - multidimensional signals
4. Signal with dependent samples
 - spectral, mean and variance change detection
 - GLR and Brandt's GLR algorithms
5. Tuning and Applications in LAbworks

Prerequisites :
Basics in Signal Processing.
Basics in Statistics and Probability.

Test

Written exam (2h) and Labworks (4h)

Calendar

S1

Additional Information

8 h CM/TD + 4 h BE + 2 h DS

Bibliography

[1] Detection of Abrupt Changes: Theory and Application, M. Basseville and I. V. Nikiforov, Englewood Cliffs, 1993
[2] Fundamentals of Statistical Signal Processing: Detection Theory, S. Kay, Prentice Hall, 1998

French State controlled diploma conferring a Master's degree

diplôme conférant grade de master contrôlé par l'Etat