AIML 429 (2024) - Home Page

AIML 429: Probabilistic Machine Learning.

Lectures and tutorials in T1, 2024:

• Tuesdays and Thursdays, 10am in Murphy LT102, Kelburn

(Videos available via "Nuku" on left side-bar, but I'd prefer you show up in class)

• Tutorials are Tuesdays from 4pm in Kirk204

Because the recordings were very poor (both sound and vision) before the room got "fixed", here are the corresponding lectures (first 3 weeks) as they were presented in 2023:

• Week 1: Lecture 1, Lecture 2 (the audio gets better after the 1st one)
• Week 2: Lecture 1, Lecture 2
• Week 3: Lecture 1, Lecture 2

Marcus's Office Hours are Tuesdays from 3-4pm , in CO337 (i.e. just prior to Tutorial -- if it's appropriate, consider asking your question in the tut so that others can hear the answer too). In-person is best. There's also zoom by arrangement.

Useful Texts

• The Algorithms For Decision Making, by Kochenderfer, Wheeler and Wray, 2023
• Murphy I, Probabilistic Machine Learning: Introduction, by Kevin Murphy, 2022
• Murphy II, Probabilistic Machine Learning: Advanced Topics, by Kevin Murphy, 2023

Materials

NotesAndQueries.pdf	these are the lecture notes, as a single document. May change as the course proceeds.
PortfolioTasks.pdf	these are the portfolio tasks. Datasets: Meal-Shirt-Wine.csv and prey-pet-temp.csv
Notebooks	Some ipython notebooks get used as demonstrations in class.

Notes and Queries is the base material for lectures. Portfolio Tasks is the set of assignment questions that you can collate. There's also a diff.pdf version if you care about seeing the `changes` made to the starting version.

Updated assessment structure:

item #	due	worth
Portfolio Part 1	Friday 19th April (end of week 6)	25%
10-min in-person Q&A	mutual arrangement in weeks 7-8 (22 April - 3 May)	20%
Portfolio Part 2	Friday 31st May, the last day of lectures	25%
1 hour test	in the exam period (date t.b.a.)	30%