Studio hour·45 minutes

Studio Hour: Split & Leakage Audit

Turn one suspicious validation result into a timeline-aware split and leakage audit.

Scenario

A baseline score is unusually high because preprocessing and feature availability were never checked against prediction time.

You will demonstrate

Draw the prediction timeline.
Identify split contamination.
Publish a reproducible correction.

Project evidence

Show the work, not a checked box.

Each response is stored in this browser as you type. Include metrics, test output, or a decision rationale wherever the deliverable asks for it.

Publish the leakage audit

Record prediction time, feature availability, row grouping, split rule, learned preprocessing boundary, before/after score, and one regression check.

Required evidence: A one-page audit table and a bounded decision about whether the original score can be trusted.

Your evidence

0/80 minimum characters

Runnable Python lab

Fit preprocessing on training rows

Calculate a training-only replacement value and apply it without reading held-out values.

Console

Run the code to inspect its output. Use Check solution only when you want to run the exercise checks.

Code runs only in a time-limited browser worker. It is not sent to the IntuitiveAI server. Do not paste secrets.

Expected result

replacement: 4.0
held out: [100.0, 4.0]
ALL TESTS PASSED

Project defense

Where should an imputer learn its replacement values?

Rubric self-review

Rate the evidence, not your effort: 0 missing, 1 weak, 2 adequate, 3 strong. All criteria must be reviewed, but a low honest score does not get hidden.

The audit connects time, grouping, preprocessing, score change, and a regression test.

Useful references

scikit-learn common pitfallsOfficial examples of inconsistent preprocessing and data leakage.

Project completion gate

Completion is controlled by stored evidence, deterministic tests, a decision defense, rubric review, and the artifact when required.

Evidence pendingCode pendingDefense pendingRubric pending

Optional cloud portfolio

Submit evidence across devices.

An account is required. Submit only when the local completion gate passes. AI review is advisory and separate from deterministic completion.

After submission, I may choose to send the stored evidence text to the configured AI model for advisory rubric feedback. Artifact bytes are not sent by this review action.

Account settings

Back to Topic Library