At bibringe deltagerne viden om statistiske metoder der benyttes til at vurdere og klassificere data. Deltagerne får fornøden baggrund til at udføre forskning inden for procesovervågning, fejldiagnose og prognose, maskin læring, objekt detektering og fortolkning baseret på store datamængder. En studerende, der fuldt ud har opfyldt kursets mål, vil kunne:

• Understand linear and logistical regression
• Understand bootstrap, maximum likelihood and bayesian methods
• Understand fitting with neural networks
• Able to work with clustering and support vector methods
• Understand supervised and unsupervised learning
• Understand hypothesis testing
• Able to work with Gaussian and non-gaussian tests
• Understand methods to design tests based on detection and false alarm probabilities

Statistiske metoder inden for maskin læring:
- linear and logical regression
- bootstrap, maximum likelihood and bayesian methods
- fitting with neural methods
- clustering and support vector methods
- supervised and unsupervised learning
- hypothesis testing
- gaussian and non-gaussian tests
- design methods based of performance (probability) T. Hastie, R. Tibshirani and J. Friedmann: Elements of Statistical Learning - Data Mining, Inference, and Prediction, Second Edition. Springer 2009.

S. M. Kay: Fundamentals of Statistical Signal Processing - volume 2: Detection theory.
Prentice Hall,1998. The contents of the course has been changed from 100% statistical change detection to now be 70% methods that form the basis for machine learning and 30% statistical change detection. 26. maj, 2020