Accurate crop yield prediction is critical for ensuring food security, optimizing resource allocation, and supporting sustainable agricultural policy development. This competition challenges participants to develop machine learning models that can accurately predict maize yield (measured in $\text{kg/ha}$) using a variety of environmental, soil, and remote sensing data sources.

Participants will work with a real-world dataset comprising field-level agricultural observations gathered from diverse geographical zones and cropping seasons. Each data point includes:

* Soil characteristics such as $\text{pH}$, organic matter, and sand/silt/clay ratios
* Daily and seasonal weather features like rainfall, temperature ($^\circ \text{C}$), and humidity
* Vegetation indices derived from satellite imagery including NDVI, EVI, and LAI
* Optional management practices like planting date and fertilizer usage
* Yield values serving as the regression target (in $\text{kg/ha}$)

The objective is to train a predictive model that generalizes well to unseen samples and minimizes error on the test set. The official evaluation metric for this competition is **Root Mean Squared Error (RMSE)**:

$$
\text{RMSE} = \sqrt{ \frac{1}{n} \sum_{i=1}^{n} \left( y\_i - \hat{y}\_i \right)^2 }
$$

where:

* $y\_i$ represents the actual observed yield
* $\hat{y}\_i$ is the yield predicted by your model
* $n$ is the number of observations in the test set

Accurate yield prediction has the potential to improve harvest planning, reduce resource waste, and enable better responses to environmental challenges—contributing meaningfully to global agricultural sustainability.