COMP4670/8600 Statistical Machine Learning

This is a broad but thorough intermediate level course of statistical machine learning, emphasising the mathematical, statistical, and computational aspects.

View My GitHub Profile

Frequently asked questions about SML

Enrolling

  1. Q: Can you give a permission code to enroll? A: We (instructors or course staff) cannot issue permission codes, you will need to fill in the online application at the college website, non-obvious cases will come to us. https://cecc.anu.edu.au/current-students/policies-and-resources/enrolling-cecs-courses

  2. Q I don’t satisfy the ANU pre-requisite requirements but I would still like to enroll (and I have evidence showing that I can handle this course). A: See response to item 1. You will need fill in the above form attaching the evidence. Such as: academic transcript, syllabus of related courses from another university, narratives about related programming or math experience.

  3. Q: What is the difference between SML (COMP4670/8600) and IML (COMP3670/6670)? Which one should I take? A: IML is a pre-requisite of SML. IML uses the “MML” book whereas SML uses the “PRML” book. Both courses cover ~ 50%+ of the chapters. Have a look at the books to get a feel of the level of the class.

  4. Q: What are the topical focus of SML (COMP4670/8600) different from that of AML (COMP4680/8650)? A: SML is a broad-but-thorough course that aims to cover most fundational topics in machine learning, whereas AML specialises in one or more sub-topics in machine learning. In recent years AML has covered covex optimisation, optimisation in ML, and other topics.

Studying

  1. Q: What programming language do we use in SML? A: Python 3.x, with both jupyter notebook and .py files.

Other

  1. Q: What if I have a question not answered on this page? A: Please file a github issue and we will address it or update this page. https://github.com/sml-anu/sml-anu.github.io/issues
  2. Q: What if I don’t like the answer of a question, or have a better answer? A: Please file a github issue (see prev item) or a pull request. Thank you for your contribution.