Broader course for all
Statistics for Physicists
Philipp Eller (TUM)
Abstract
This course introduces the core statistical concepts needed to carry out data analyses in physics. We begin with the fundamentals of probability theory, so no prior exposure to statistics is required. Building on this foundation, we will explore the two major paradigms of statistical inference: Frequentist and Bayesian.
On the Frequentist side, we will learn how to use data to estimate physical quantities, study the properties of estimators, and develop a general framework based on Maximum Likelihood Estimation (MLE). We will then turn to hypothesis testing, where we will revisit the classic Neyman–Pearson lemma, followed by methods for constructing confidence intervals.
In the final part of the course, we shift to Bayesian inference, a conceptually different approach that introduces its own techniques and tools. Starting from Bayes’ theorem, we will see how prior information and data combine to yield posterior distributions for parameter inference and model comparison. We will also discuss the conceptual contrasts between Bayesian and Frequentist thinking, highlighting their differences and points of agreement.
After completing this course, you will have gained the foundational knowledge and practical skills required to embark on your own data analyses in astro- and particle physics.
