ECTS credits:  5

Course parameters:

Language:  English

Level of course:  PhD course (can be joined by master project students)

Time of year: autumn semester 2026

No. of contact hours/hours in total incl. preparation, assignment(s) or the like:  3/9 hours per week

Capacity limits:  none

Course fee: none for Danish PhD students

 

Objectives of the course:

The course gives an overview of statistical methods used in data treatment. The purpose is to give a better understanding of the limitations in standard data analysis procedures and to introduce more advanced techniques used e.g. for limited counting statistics. The course is addressed at second year master students or PhD students and the techniques employed will be relevant mainly for experimental nuclear-, particle- and astro-physics as well as observational astronomy.

 

Learning outcomes and competences:

At the end of the course, the student should be able to:

• Explain fundamental concepts in statistical data analysis.
• Judge and apply methods for estimation of parameters, for setting confidence limits and for goodness-of-fit testing.
• Discuss how systematic uncertainties are handled.
• Compare the pros and cons of using classical versus Bayesian methods.
• Explain the principles behind robust statistical methods.

 

Compulsory programme:

One mandatory assignment

 

Course contents:

The course builds upon the introductory courses in data treatment and statistics. The theory of estimation is covered with emphasis on maximum likelihood and least squares methods. The principles behind confidence-interval setting are treated in detail and several methods for goodness-of-fit tests are discussed. Robust methods and EDF (empirical distribution function) methods are introduced and differences between classical and Bayesian methods discussed. Methods to handle systematic uncertainties are discussed. The relation to machine learning techniques will be stressed. The starting point will, as far as possible, be the participants’ concrete needs and problems in data analysis. Numerical implementations are introduced if needed

 

Prerequisites:

Previous experience in data analysis

 

Name of lecturer[s]:

Karsten Riisager

 

Type of course/teaching methods:

Lectures (2h/week) and exercises (1h/week)

 

Literature:

'Data Analysis in High Energy Physics' eds O. Behnke, K. Kröninger, G. Schott and T. Schömer-Sadenius

 

Course homepage:

Brightspace

 

Course assessment:

Oral presentation of the mandatory assignment

 

Provider:

Department of Physics and Astronomy

 

Special comments on this course:

 

Time:

To be determined

 

Place:

To be determined

 

Course fee:

none

 

Registration:

For registration: via e-mail to kvr@phys.au.dk, at latest on August 17, 2026

If you have any questions, please contact Karsten Riisager, e-mail: kvr@phys.au.dk