The goal of this course is to introduce students to basic algorithms for learning from example data, focusing on classification and clustering problems. This course is intended for 3rd/4th/5th year undergraduates and MSc students. It is delivered using an
Inverted Classrom method. It should NOT be taken with MLPR, which covers much of the same material.
The syllabus for this course is defined by the lecture slides, the tutorial questions and the labs. Any items from any of those sessions may be examined on. The required understanding of a subject may go beyond what is simply presented in the slides: you should know how to apply it and the more detailed context. The course texts and the videos can help with that.
Class Meetings
IMPORTANT INFORMATION: This course is not taught in the traditional lecture style. The expectation is that you take more control of your education in this course. This means that you will have about 20 hours of video to watch in your own time. This material is assessable, except where noted. It is very important that you watch the videos and do the associated quiz for the class meeting topics at least a day before the class . Each class will consist of several parts: 1) discussing any questions about the videos that you either suggest in advance or raise in class on the day, 2) simple non-assessed exercises to explore the issues raised in the videos that you have just watched, 3) exercises or examples based on quiz questions that were difficult, and 4) examples going further into depth in specific subtopics.
Topic Content
Each topic listed in the Topics section has a set of subtopics in a particular order, and it will make most sense if you go through them in this order. For each subtopic, there are the following resources:
- PDFs of the slides used in the video for the subtopic.
- Video for the subtopic
For the topic as a whole, there is also the following:
- Consolidated PDF of all the subtopic slides.
- Video playlist for all the subtopic videos - this can be easier if you want to watch all the videos for a topic in one session.
- In some cases, links to external resources to provide additional information and/or other perspectives on the topic.
- A self-assessment quiz for topic content - you may take this multiple times and it is intended for you to assess how well you have learned and understood the topic content.
Other Resources
Past years' exam papers are available
online. Solutions are not available.
The
lecture notes from the old Learning from Data course are useful, although they contain more mathematical detail than we are expecting for IAML.
Piazza Q&A
Piazza is the place to ask questions about the course materials: topics slides and videos, the labs, tutorials and the assignments. If you do not already have access to the "Introductory Applied Machine Learning (2021-2022)[SEM1]" piazza page, you can sign up by clicking the "Discussion (Piazza)" link on the left of this page (this should be visible if you are enrolled in the course).
Some key points to keep in mind:
- Before posting anything, check that your question has not already been answered previously. Searching through the topic-folders is a great way to find other similar questions with relevant answers.
- For the same reason above and to help us to address your question in a timely manner, tag your query/comments with the appropriate topic label: for example 'labs' if it is an issue with a lab question, or 'logistics' if administrative issue or gradescope signup problems.
- We encourage students to answer questions if you can - it is a great learning experience to explain something to another student.
- If you are uncomfortable asking a question publicly you have several options:
- You can post questions anonymously to other students: Note however that TA's and Instructors will be able to see your identity.
- You can post private questions which can be seen only by Instructors/TA's. This is the only method allowed when asking specific questions about the assignments.
- If your issue should be kept confidential (i.e. not seen by TA's), then of course please email the course lecturer.
Coursework
There will be one assessed coursework, using Python and key libraries used in machine learning. The coursework is worth 30% of the overall course mark.
Note that the Labs are strongly advised as preparation for the coursework as each Lab introduces you to the libraries.
See the Labs & Assessment pages for details.
Tutorials
Tutorials will be in weeks 3, 5, 7 and 9. You will be allocated to a tutorial slot in your timetable, please stick to this group throughout the semester. If you cannot make the time then you can request a change via the
Group Change Request Form. See the Tutorials page for more information.
Labs
Assignment-related labs will be in weeks 2, 4, 6, 8, and 10 with an optional introduction to Python and packages in week 1. You will be allocated to a lab group. If you cannot make the time then you can request a change via the
Group Change Request Form. See the Labs page for more information.
Exam
The exam, worth 70% of the course mark, will be in the December diet. See the Assessment page.
Enrolling in IAML
If you can see the Labs link on the left menu, you are already enrolled in IAML!
To be included in Lab and Tutorial groups you need to be enrolled for credit in Euclid. If you want to take IAML for credit, you must get your personal tutor or supervisor to enrol you in Euclid, which will then automatically enrol you here on Learn. IAML is not available to visting undergraduate students.
