Mine Basket History to Justify Grocery Shelf-Adjacency Changes
Overview
What this challenge is about.
Mine basket data with FP-growth, filter high-lift pairs, and rank shelf-adjacency changes by difficulty to earn a verifiable certificate.
The scenario
Maison Coteau is a Lyon-based regional grocery chain whose 180 stores generate dense basket-level scan data, yet its category and merchandising teams still set shelf layouts by habit and supplier pressure rather than by what customers actually buy together.
The Brief
What you'll do, and what you'll demonstrate.
Determine which product pairs sell together strongly and reliably enough to justify changing where they sit on the shelf, and prove the expected payoff before committing to a rollout.
Earning criteria — what you'll demonstrate
- Tune frequent-itemset mining parameters to the differing purchase frequencies of distinct product categories
- Distinguish genuine, durable association from promotion-driven coincidence using lift and time-stability checks
- Translate statistical patterns into ranked, implementable business recommendations
- Design a statistically defensible retail A/B pilot to validate an expected effect before rollout
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Aligned coursework coming soon.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
- Frequent Itemset Mining
Apply frequent itemset mining to solve real industry problems and demonstrate production-level capability.
- Fp Growth
Apply fp growth to solve real industry problems and demonstrate production-level capability.
- Spark
Apply spark to solve real industry problems and demonstrate production-level capability.
- Lift Analysis
Apply lift analysis to solve real industry problems and demonstrate production-level capability.
- Ab Testing
Apply ab testing to solve real industry problems and demonstrate production-level capability.
- Data Storytelling
Apply data storytelling to solve real industry problems and demonstrate production-level capability.
Careers
Career paths this challenge builds toward
Completing this challenge demonstrates skills that transfer directly to these roles:
Data Engineer (Retail Analytics)
This challenge mirrors a core retail-analytics task: turning large, messy transaction logs into trustworthy patterns. You practice tuning a mining algorithm to real data characteristics and building a reproducible pipeline that downstream teams can rely on for layout decisions.
This challenge sharpens
- frequent-itemset-mining
- fp-growth
- spark
Decision Scientist
You move from raw association statistics to a defensible business decision, separating real signal from promotional noise and validating expected impact with a pilot. That bridge from analysis to evidence-backed recommendation is the heart of a decision-science role.
This challenge sharpens
- lift-analysis
- ab-testing
- data-storytelling
Experimentation Engineer
Designing a statistically sound store-level A/B pilot, sizing it from a target effect, and guarding against confounds like promotions is exactly what experimentation engineers do to validate merchandising changes before broad rollout.
This challenge sharpens
- ab-testing
- lift-analysis
- fp-growth