01 / STARTER
Leakage-aware classification
Document the split and fit preprocessing only on training data. Compare the model with a baseline.
PATHWAY / MACHINE LEARNING
Show what your model gets right. And wrong.
Machine learning portfolio projects with baselines, leakage checks and reproducible evaluation.
Three mission briefs
A suggested progression, not a certification track. The briefs are prompts; reference builds are not included.
02 / NEXT
Small tabular baseline model
Use a licensed or synthetic dataset and a simple baseline to define what improvement would mean.
03 / NEXT
Model evaluation and model card
Repeat the evaluation, inspect errors and explain where the model should not be used.
What useful evidence looks like
Show the method and leave room for uncertainty.
The short version
I compared a simple baseline with a classifier using the same holdout set. Preprocessing was fitted on training data. I recorded errors and the evaluation settings. The experiment is small and has not been validated on real deployment data.
Illustrative summary. No completed result or learner identity is claimed.
Bring these checks
- Data source, permission, target and relevant exclusions are documented.
- The split rationale handles time or repeated entities where relevant.
- Preprocessing is fitted using training data only.
- A simple baseline and suitable metrics use the same evaluation setup.
Keep going
Write down the result before you package it.
GUIDE
Model card for a portfolio project
A practical structure for explaining your project and its limits.
TOOL
Evidence checklist
Review the requirements for your pathway without an overall skill score.
TOOL
Evidence-pack builder
Turn your notes into a locally downloadable project summary.