VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.

402-0738-00L 10 Credits BSC , DZ , SHE , MSC D-GESS , D-PHYS , D-MATH
You're viewing possible stale or outdated data. Please check the latest semester for more up-to-date information.

Statistical Methods and Analysis Techniques in Experimental Physics

Lecturers & Examiners: Prof. Dr. Mauro Donegà
VVZ CR n/a

Last Updated: 2026-06-01 11:33:33

Abstract

This lecture gives an introduction to the statistical methods and the various analysis techniques applied in experimental particle physics. The exercises treat problems of general statistical topics; they also include hands-on analysis projects, where students perform independent analyses on their computer, based on real data from actual particle physics experiments.

Objective

Students will learn the most important statistical methods used in experimental particle physics. They will acquire the necessary skills to analyse large data records in a statistically correct manner. Learning how to present scientific results in a professional manner and how to discuss them.

Content

Topics include: - Probability and probability distributions -- - moments, quantiles, covariance -- - combinatorial calculus reminders -- - most used continuous/discrete probability distributions - Measurement uncertainties -- - statistical / systematics uncertainties -- - error propagation - Monte Carlo methods -- - random numbers -- - hit/miss -- - inverse cumulative - Parameters estimation -- - likelihood and least squares fits - Introduction to Bayesian statistics -- - MCMC diagnostics -- - bayesian Evidence -- - simulation-based inference -- - ABC -- - advanced sampling algorithms - Hypothesis testing -- - goodness of fit -- - two samples problem -- - resampling techniques - Confidence intervals -- - confidence belt classical construction (double sided, upper/lower limits) -- - Limits near boundaries / Feldman-Cousins -- - LHC test statistics / asymptotic formulas - Machine Learning (Multivariate Analysis Methods) -- - Fisher discriminant -- - Boosted decision trees -- - Neural Networks - Unfolding techniques -- - matrix inversion -- - regularisation -- - fit approach Methodology: - lectures about the statistics topics; - common discussions of examples; - exercises: specific exercises to practise the topics of the lectures; - all students perform exercises with python/notebooks on their laptops; - students complete a full data analysis in teams (of two) over the second half of the course, using real data taken from particle physics / astroparticle / cosmology experiments; - at the end of the course, the students present their analysis results in a scientific presentation; - all students are directly tutored by assistants in the classroom.

Resources

Lecture Notes

- Copies of all lectures are available on the web-site of the course.- An interactive jupyter-book scriptum of the lectures is also available to all students of the course.

Literature

1) Introduction to error analysis, J.R. Taylor, University Science books: ISBN-10: 093570275X 2) Statistics for nuclear and particle physics, L.Lyons, Cambridge University Press; ISBN-10: 0521379342 3) Statistics: A guide to the use of statistical methods in the Physical Sciences, R.J.Barlow 4) Statistical data analysis, G. Cowan, Oxford University Press; ISBN-10: 0198501552

General Information

Language
English
Levels
BSC , DZ , SHE , MSC
Frequency
Yearly recurring

Examination

Type
session examination
Mode
oral 30 minutes
Based on the regulation of continuous performance assessments, students have to work on a project and present the results at the end of the semester. This additional task is graded on a pass/fail basis and students who fail this task have to repeat the complete course.

Course Components

Type Title Time & Place Hours
lecture with exercise Statistical Methods and Analysis Techniques in Experimental Physics
  • Tue 08:45-13:30 (HIT F 21)
5 h weekly

Offered In