Help
Bring the smallest example that demonstrates the problem. Keep the people in it fictional.
I am stuck on an exercise#
Open the week’s lecture and tutorial together. The lecture introduces the model; the tutorial specifies what to produce. Check the named input artefact before debugging the output.
For a calculation or code problem, bring:
- The question you were trying to answer and your stated assumptions.
- The smallest synthetic input that reproduces the problem, plus a random seed if sampling is involved.
- Your expected result, actual result and the step where they first differ.
“My rank is low” is an observation. “Changing clarity from 2 to 3 reduced my score” is a useful test case.
I need the files or calculator#
The course toolkit contains the downloads and worked examples. Its A/B calculator runs in your browser; it does not upload your inputs. Without JavaScript, the worked example and dataset links remain available. Download the reference scorer with the synthetic controls so your project can reproduce its benchmark offline.
I need to speak to someone#
In this fictional course, Casimir Beng holds consultation on Mondays, 12–1 pm, in Systems Lab 2, after the lecture. Bring questions about assumptions, assessment, extensions or a deferred exam.
Joost Nwosu is available on Thursdays in weeks 2–12, 3.30–4 pm, in Systems Lab 2, after the tutorial, for exercises and reproducibility. These describe the invented teaching arrangements, not a live support service.
For access adjustments, submission arrangements or an interrupted assessment, read the course policies and raise the issue at consultation. You never need to disclose a relationship history to explain a learning need.
I am not sure what the course expects#
The FAQ covers tools, workload and the synthetic-profile benchmark. The assignment briefs are the source of truth for deliverables and marking; the timetable puts them alongside the teaching weeks.