IndabaX 2026 Emission Forecasting Competition – Nigeria
Helping Nigeria
TBD
Under code review
Prediction
73 joined
49 active
Start
May 02, 26
Close
May 14, 26
Reveal
May 14, 26
About
The dataset includes both input features adn the target variable for prediction. Below is a detailed description of each variable:
Variable Type Description
ID String Unique 10-character identifier for each observation.
Latitude Float Latitude coordinate of the measurement location (in decimal degrees)
Longitude Float Longitude coordinate of the measurement location (in decimal degrees)
DayOfYear Integer Day of the year when the measurement was taken (1 - 366). Helps capture seasonal patterns
DayOfWeek Integer Day of the week when the measurement was taken (1 = Monday, 7 = Sunday). Useful for weekly cycles
Hour Integer Hour of the day when the measurement was taken (0 - 23). Captures daily patterns.
Month Integer Month of the year when the measurement was taken (1-12). Useful for seasonal trends
Emission Float Measured emission level at the given location and time. This is the target variable for prediction.
- Temporal and spatial features: Use DayOfYear, DayOfWeek, Hour, Month, Latitude, and Longitude effectively for modelling trends over time and across locations.
- Target variable: Emission is continuous, consider regression models suitable for continuous prediction.
- Avoid leakage: Do not use future data from the test set for training.
- Leaderboard strategy: Optimise for the public set, but do not overfit - the private set determines the final ranking.
Join the largest network for
data scientists and AI builders
Status
© Zindi 2026