In the project, you will use machine-learning techniques in a realistic setting. We will provide a set of projects and corresponding datasets that you can choose from. You will work in groups of 4 people¹, using Piazza to find teammates, and will collaborate to produce a project report that will be assessed.

Each group only submits one report. All members of the group will receive the same grade, excepting any special/extenuating circumstances. The grade will be based on the final report only.

You will have considerable freedom in the projects, but it should involve most parts of techniques described in the lectures. This would typically involve

• reading up on some relevant background to well understand the task and what has been done previously (via google scholar, internet search, in some cases references are provided)
• some exploratory data analysis
• if classification is the goal, choosing methods that might work well on the task, based on the earlier steps
• evaluating the results of the different methods on the task (e.g. assessing generalisation performance).

Note that you’re not required to outperform prior methods. The important thing is that the approaches taken are reasonable, methodologically correct, and clearly described in the report. Good projects would nonetheless discuss possible reasons for performance differences compared to more advanced approaches.

• Sample project reports that provide a rough guide for what goes into the projects
  • Example 1: scored around 70% and slightly above
  • Example 2: scored around 60%

Note: these reports are 8 pages long, but your reports only need to be 6 pages.

FAQ

1. Are we allowed to explore methods that are not explicitly covered in the course?

   Yes, that’s perfectly fine. There is no requirement to stick to just using what is taught in the course. Having said that, it is your responsibility to ensure that your report covers enough of the background material for whatever methods you employ so that an intelligent reader will be able to follow your reasoning and evaluate your findings.

2. What needs to be put in the main report and what goes in the appendix?

   All the text and figures required to understand and evaluate your contributions need to go within the 6-page limit. References and statements of contribution for team members are not counted in the page limit.

   You may also optionally include an appendix with additional information, results, or examples. These are not directly evaluated; much like is standard practice for academic papers. The purpose of an appendix is to allow for extra material that is not essential to the understanding of the main paper, but helps provide a more comprehensive understanding of the undertaken research. The reviewers/markers are not obligated to read appendices/supplementary material. They may choose to do so, but don’t have to.

3. What details about the team should we provide in the report.

   You should only provide the student IDs (UUNs) of the team members (not exam numbers). Please don’t provide your names, emails, or other information. The LaTeX template reflects this.

4. What should we provide in the statement of contributions at the end of the report?

   You should provide a brief description of what each team member did for the project. This can be as simple as 2-3 sentences for each individual outlining the tasks the performed, e.g. implementing the methods in Section X, writing section, Y, jointly coming up with the main ideas, co-writing Section Z, etc. Do not use your names; use only your student ID number.

5. Can we use Generative AI (ChatGPT, etc) for the coursework?

   Please see Guidance on the use of Generative AI in the course information page.

Important Dates

Fri 03 Oct, 12pm

Each group (any one member) should fill-in the project details form with the student numbers (UUNs) of the members and the choice of project topic.

Wed 29 Oct, 5pm

Each group (any one member) must upload a progress report using the project progress report form. This is not assessed. The goal is to ensure your project has the right scope and that you are on track. The report must be at least two pages, in the given latex template, and can include whatever you’ve already written up thus far. To help us understand your plans, please include brief sections on “Current Progress” and “Plans for Completion” beyond anything already written up.

Fri 31 Oct, 1-4pm AT 5.04

When submitting the progress report, you may choose to come to a drop-in session where we can provide some basic feedback. It is entirely optional, and attendance is not enforced. NOTE This session is primarily to help those who may have potentially gone off track or fallen behind, realign and focus on finishing up things.

Thu 20 Nov, 12pm

Final report due. You should submit the report as a single PDF, including references and appendices, with the filename as your 2-digit group number. For example, group 6 would submit 06.pdf.

The maximum page limit is 6 pages, and the references, statement of contribution, and appendices do not count towards the page limit.

Authors’ Instructions

The report should be maximally 6 pages long (excluding references), using this report template² (adapted from the NeurIPS template). It should contain the following in some manner:

• description of the task
• relevant background and related previous work
• explanation of the significance/relevance of the objective/task
• information on the data preparation
• exploratory data analysis
• description of the learning (e.g. classification) methods used
• results and evaluation
• conclusions

You are expected to discuss the work within your group, and to work on your report together. You should write up the project as a whole, including the work of the others in your project. Please cite your sources (data, methods, etc.) appropriately.
At the end of the report, there should be a short description of how each member of the group contributed to the project, which can be on an additional page (not counted for page limit).

Marking

The marking criteria include the appropriateness of the machine learning methods chosen, quality of the analysis, the quality of the evaluation, the amount of work, and the quality of the explanation of the report (both text and graphics)—in keeping with the components listed above for the report. While you will be marked out of 100 in line with the common marking scheme, an interpretation of the scheme can be seen as:

70-100

Excellent explanation and description of points above plus extra achievement at understanding or analysis of results. Clear explanations, evidence of creative or deeper thought will contribute to a higher grade.

60-69

Excellent explanation and description of points above, with some minor deficiencies.

50-59

Well explained description of points above, be some deficiencies with writing or work undertaken.

40-49

Sufficient description of points above but significant deficiencies.

30-39

Evidence that the student has gained some understanding, but not addressed the specified task properly.

0-29

Serious error or slack work.

Good Scholarly Practice

Please remember the good scholarly practice requirements of the University regarding work for credit. You can find guidance at the School’s page on academic misconduct.

---

Projects

To be announced.