Graphical Models

Course content

  • Markov kernels and conditional distributions
  • Probabilistic conditional independence
  • Conditional independence models
  • Markov properties on undirected, directed, and bidirected graphs
  • Bayesian networks
  • Local computation in junction trees
  • Gaussian graphical models
  • Estimation of graph structure
Education

MSc Programme in Statistics

Learning outcome

Knowledge:

Basic knowledge of the topics covered

Skills:

  • Understand simple properties of conditional distributions and Markov kernels
  • Discuss and understand issues concerning conditional distributions and the interplay between probabilistic and other types of conditional independence
  • Ability to use standard software packages for the analysis of simple graphical models

 

Competences:

  • Understand graph based Markov properties and their role for simplification of computation and interpretation
  • Understand properties and limitations of methods for estimating graph structure

Four hours of lectures and three hours of exercises per week for 7 weeks.

Lecture notes and selected chapters from books

Basic mathematical statistics and probability based on measure theory.
I.e. Measures and Integrals + Stat1 + Stat2 or equivalent.

ECTS
7,5 ECTS
Type of assessment
Written assignment, 27 hours
Written take-home assignment
Aid
All aids allowed
Marking scale
7-point grading scale
Censorship form
External censorship
Criteria for exam assessment

The student must in a satisfactory way demonstrate that he/she has mastered the learning outcome of the course.

Single subject courses (day)

  • Category
  • Hours
  • Lectures
  • 28
  • Theory exercises
  • 18
  • Practical exercises
  • 3
  • Preparation
  • 130
  • Exam
  • 27
  • English
  • 206