Store demand forecast

The brief

Stores run out of popular items before festivals and are left with unsold stock after. Forecast daily sales per store and product with weekly cycles, holidays and promotions, and prove the forecast beats a simple baseline. Good looks like: forecasts across many series, honest back-testing, and errors reported by store and category.

What you will build

1. Prepare daily sales series per store and item
2. Add calendar, holiday and price features
3. Train a gradient-boosted forecast for many series at once
4. Back-test against a seasonal baseline
5. Report errors by store and category

Free, open data to build it with

• M5 forecasting (Walmart) ↗ Kaggle

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.