Programme And Module Handbook
 
Course Details in 2026/27 Session


If you find any data displayed on this website that should be amended, please contact the Curriculum Management Team.

Module Title LM Inference from Scientific Data
SchoolPhysics and Astronomy
Department Physics & Astronomy
Module Code 03 23560
Module Lead Dr Chris Moore
Level Masters Level
Credits 10
Semester Semester 1
Pre-requisites
Co-requisites
Restrictions None
Contact Hours Lecture-18 hours
Guided independent study-82 hours
Total: 100 hours
Exclusions
Description

Data is precious, and the aims of any good observer or experimentalist should be to make best use of it. To do this requires an understanding of statistics. Most undergraduates regard statistics as a boring necessity.
This is entirely understandable, since it is often introduced into undergraduate laboratories in the form of a set of rules which must be followed to calculate and propagate errors. In fact there is much more to the subject than this. Much of the scientific enterprise can be regarded as one of making comparisons between models of reality and the data with which they are tested. Moreover the area is the subject of a lively debate between two schools of thought - Bayesians and frequentists - who disagree about the whole role of probability in the analysis of data. Once one understands this, the subject becomes much more interesting. The main aim of the course is to give you the beginnings of such an understanding, and to help you to apply the tools of statistical analysis in your own research problems. Although the techniques introduced are entirely general, they will be illustrated using examples drawn mostly from the lecturer's own field: astronomy. The course is assessed via a variety of practical analysis exercises which are also mostly astronomical in nature. Some interest and experience in astronomy would therefore help you to connect with these examples.

Learning Outcomes

 

  • Understand the difference between Bayesian and frequentist approaches to scientific inference.
  •  Understand the use of probabilities in scientific data analysis, and the use of Bayes' Theorem.
  • Be familiar with some of the most common probability distributions encountered when working with scientific data.
  • Be able to formulate many problems of scientific inference in a Bayesian fashion, and to understand the concepts of prior and posterior probabilities.
  • Be able to apply frequentist approaches to test for correlation, to fit models and to test for the presence of a significant signal.
Assessment 23560-01 : Examination : Exam (Centrally Timetabled) - Written Unseen (60%)
23560-03 : Coursework : Coursework (40%)
Assessment Methods & Exceptions Coursework (40%); 1.5 hour Examination (60%)
Other None
Reading List