Practice

Optional exercises — not graded

The syllabus states it plainly: no individual or group assignments are required. Nothing on this page is graded. It exists because the exam has a coding part, and the only way that part goes well is if you have typed the code before.

How it works

Each set goes live after the lecture it belongs to: 30 to 45 minutes, one dataset, a handful of questions.

  1. Do it in your own R script. Every set starts with a set.seed() — use it, so your numbers can be checked.
  2. Check yourself on Blackboard. Each set has a test with the same name (Set 10, Set 11, …). Type in your numbers; the feedback arrives the moment you submit. As many attempts as you like, and it counts for nothing: it is a mirror, not an exam.
  3. Bring the sentences. Questions marked In class end in a judgement, not a number, so no test can check them. We discuss them at the start of the lecture named under the question.

Solutions are posted one week later. Work in R unless a set says otherwise — set 16 is in Python, like the lab it follows. See Setup if your machine is not ready, and bring what breaks to office hours.

The test checks the numbers; the exam asks you to justify a choice. That is why the In class questions matter more than the score.

Module 1 · Prof. Arbia

Set After lecture Topic Exercises Solutions
1 14 Sep Hypothesis testing and confidence intervals — —
2 18 Sep OLS, prediction, bias–variance, cross-validation — —
3 25 Sep Multiple regression — —
4 29 Sep Nonlinear regression — —
5 2 Oct GLMs: logit and Poisson — —
6 5 Oct Survival models — —
7 9 Oct Ridge, LASSO, Elastic Net — —
8 16 Oct Time series components and smoothing — —
9 23 Oct AR and ARIMA — —

Module 2 · Prof. Salvini

Set After lecture Topic Exercises Solutions
10 3 Nov k-means and customer segmentation set —
11 9 Nov Hierarchical clustering set —
12 16 Nov Principal components set —
13 17 Nov Correspondence analysis set —
14 23 Nov Decision trees set —
15 27 Nov Random forests and gradient boosting set —
16 30 Nov Neural networks · Python set —
17 9 Dec Interpretability set —
18 11 Dec Dashboards and communication set —

Exam simulations

Past intermediate exams and mock papers appear here before each session.

Session Paper Solutions
Mock — Module 1 before first intermediate — —
Mock — Module 2 before second intermediate — —