Skip to content

Exams

Two summative exams bookend the second half of each module. They use the same question styles as the quizzes, scaled up: MCQ, code tracing, debugging, short answer, practical coding, and scenario/problem-solving. The practical coding part is answered directly in a Jupyter notebook during the exam.

Exam When Time Covers CLOs
Midterm Week 8 · Session 16 90 min Weeks 1–8 — the full CLO-1 workflow: Python → Jupyter → Git → NumPy → Pandas → cleaning → acquisition & combining → visualization → EDA CLO-1 (+ entry-level CLO-2)
Final Week 16 · Session 32 120 min Weeks 9–16 — machine learning (CLO-2) and AI-assisted workflows (CLO-3), including a reproducibility & AI-ethics section CLO-2, CLO-3

Structure (both exams)

  1. Part A — Written (~25–40 min): definitions, code-reading ("what does this output?"), one-liner code-writing, and explain-why questions.
  2. Part B — Practical notebook (remaining time): a small, provided dataset and a task list — load → inspect → clean (with justification) → derive → filter/sort → charts → written findings with evidence.

The midterm session page and the session 32 page describe the exact rules and timing as students experience them.

How to prepare

  • Midterm: rework Lab 16 · Midterm Review — it is a timed mock exam in the real shape. Review each session's key concepts and terminology for Sessions 1–15.
  • Final: Modules B and C content (Sessions 17–30). Be ready to explain your models, not just fit them — expected questions include "why did you choose this metric?", "what does this confusion matrix cell mean?", and short ethics/reproducibility scenarios.
  • Re-run the course notebooks (10–20) in a fresh environment so the ML and AI-tool code is familiar.
  • Practice the findings with evidence style: every claim you make in the notebook exam should have the printed output beneath it.

Exam-day rules

  • Written part: closed notes. Practical part: official pandas / NumPy / scikit-learn reference docs allowed; saved notebooks and AI assistants are not — these exams assess individual skill (CLO-3 AI skills are assessed in Sessions 25–30, labs, and the final project instead).
  • Restart & Run All your notebook before submitting, then push to the exam branch and note the commit hash.
  • Justify every cleaning choice in one sentence — partial credit exists for reasoning, and unexplained dropna() calls lose marks.