All projects
OngoingApplied AI/MLApplied AI/ML · Time-Series Forecasting
Forecasting Nigerian Food Prices
Exploring forecasting methods for staple food prices in Nigeria, from seasonal baselines to gradient boosting and neural forecasters.
Objective
Explore forecasting methods for Nigerian food prices.
Methods planned
- Data cleaning.
- Lag features.
- Rolling statistics.
- Seasonal naive forecasting.
- ARIMA/SARIMA.
- Gradient boosting.
- Neural network forecasting.
- Rolling-origin backtesting.
Data sources
- World Food Programme food-price data (potential source)
- National Bureau of Statistics, Nigeria (potential source)
Scope
- Rice
- Beans
- Garri
- Yam