PML: Probabilistic Machine Learning

Lecturer: Michael Gutmann  
Teaching assistant: Hayden Ramm (contact: h.r.ramm  at sms.ed.ac.uk)
Lecture/tutorial times and locations: please see here.

Content

The course covers foundational material in probabilistic machine learning with a focus on unsupervised learning. It will provide you with tools and skills to understand many different methods from first principles and develop new ones. The course is organised around five main topics:

1. representing probabilistic models
2. exact inference
3. actions and decision making
4. learning
5. approximate inference and learning

The course material is available on these pages. The lecture recordings are available here and via Learn. We use EdStem for course Q&A (access via Learn).

Assessment

  • The course is assessed by a written exam at the end of the course (100% of final mark). There is no assignment.
  • The exam date is announced here and past exams are available here (search for probabilistic modelling and reasoning, course code: INFR11134).
  • PML follows the Common Marking Scheme of the university and the school-wide academic conduct policy.

Formative quizzes

The course has three formative (not assessed) quizzes on the following topics:

  • Directed and undirected graphical models (Quiz 1). It covers the material on the slides until (and including) “Undirected Graphical Models II”, and the exercises discussed in tutorials 1 and 2.
  • Inference and message passing (Quiz 2). It covers the material on the slides “Exact Inference”, the inference self-study questions, and the inference exercises discussed in tutorials 3 and 4. 
  • Learning (Quiz 3). It covers the slides “Basics of Model-Based Learning”, “Factor Analysis and Independent Component Analysis”, "Intractable Likelihood Functions", the corresponding self-study exercises, and Tutorial 6. 

The quizzes are available on Gradescope that you can access it via Learn

Please note that the quizzes are designed to support your engagement with the course and help you assess your understanding as you progress. They are intentionally easier than the exam.

License
All rights reserved The University of Edinburgh