Sprints
Sprints split the project into stages that follow the course content. They are not graded: they help track progress and give you feedback in time to act on it.
How they work
- From Sprint 1 on, your project notebook is shared with hizocar@gmail.com (commenter access). You don't send anything for each sprint: I review the current version in Colab.
- Reviews take place during the workshop of the session listed below. Each team gets a traffic light:
| Traffic light | Meaning |
|---|---|
| 🟢 Green | On track. Keep going. |
| 🟡 Yellow | Something needs adjusting before the next sprint. Check the comments in your notebook. |
| 🔴 Red | There is a problem that could affect the final submission. Let's talk during the same workshop. |
| Sprint | Content it applies | Starts in | Reviewed in |
|---|---|---|---|
| 1 · Data in Colab | Session 1: Python and Colab | Session 1 | Session 2 (Nov 4 and 6) |
| 2 · Data cleaning and exploratory analysis | Session 2: pandas and seaborn | Session 2 | Session 3 (Nov 11 and 13) |
| 3 · First model | Sessions 3 and 4: regression and classification | Session 3 | Session 5 (Nov 25 and 27) |
| Final submission | Session 5: trees and communication | Session 5 | Dec 1 (submission), Dec 2 and 4 (presentation) |
Sprint 1 · Data in Colab
Worked on in the session 1 workshop. Reviewed in session 2.
- Team registered (2 or 3 members) and case chosen. If it is your own case, include the written proposal for approval.
- Notebook created in Colab and shared with the whole team and with hizocar@gmail.com.
- The data is loaded from a URL with
pd.read_csv. - First inspection:
shape, data types, first rows, number of missing values, and number of duplicates. - The case's business question written in your own words in two or three lines.
Sprint 2 · Data cleaning and exploratory analysis
Worked on in the session 2 workshop. Reviewed in session 3.
- Duplicates removed and missing values handled, with the decision justified.
- Inconsistent categories unified (for example,
"bachelor","Bachelor", and"BACHELOR"). - Impossible values (negative ages, scores out of range) detected and fixed.
- Cleaning summary table: issue, decision, and rows affected.
- At least four exploratory charts, including how the variables relate to the target.
- Three written findings, each with what it means for the business.
Sprint 3 · First model
Starts in session 3 and continues in the session 4 clinic. Reviewed in session 5.
- Categorical variables encoded (for example, with
pd.get_dummies). - Data split into training and test sets before training.
- Baseline model trained: linear regression (case A) or logistic regression (case B).
- Metrics computed on the test data and chosen to fit the problem.
- At least two coefficients interpreted in business terms.
Final submission
Worked on in the session 5 workshop.
- Second model (decision tree or random forest) trained and evaluated on the same split.
- Model comparison table and a justified choice.
- Feature importance for the chosen model.
- Three recommendations and the limitations of the analysis.
- A 10-minute presentation with a pyramid structure.
- Checklist from the Colab guide completed.
All submission details and the rubric are in the project brief.