Primary competition visual

2025 DAUST Hackathon by DAUST

Helping Africa
500 Points
Closing soon! (9 days left)
Prediction
Earth Observation
2 joined
0 active
Starti
Dec 12, 25
Closei
Jan 09, 26
Reveali
Jan 10, 26
About

You’ll be working with a dataset compiled from African farms, containing:

  • Measured soil chemical properties (in ppm)
  • Environmental and spatial features
  • Bulk density and other agronomic indicators

The data reflects soil conditions at a depth of 20 cm, which is standard for assessing nutrient availability for most crops.

Nutrient Gap Calculation

Soil test values (in ppm) are converted to available nutrients per hectare (kg/ha) using the formula:

Available (kg/ha) = ppm × soil depth (cm) × bulk density (g/cm³) × 0.1

The required nutrients for a 4-ton/ha maize yield are calculated as:

Required (kg/ha) = Nutrient uptake per ton × Target yield (4 t/ha)

Then, the nutrient gap is computed as:

Gap (kg/ha) = Required − Available

Nutrient requirements are based on empirical agronomic data such as:

  • Nutrient omission trials
  • Plant tissue analysis
  • Crop simulation studies

Example: If maize yields 4 t/ha and total nitrogen uptake is 100 kg/ha, the N requirement per ton is 25 kg/t, leading to a total required N of 100 kg/ha.

If a value is negative it means there is excess of that nutrient in the soil already and the farmer does not need to add any more. If the value is positive then the farmer needs to add those nutrients to the soil.

Nothing to add.

Files
Description
Files
Additional earth observation data.
Additional earth observation data.
Additional earth observation data.
Is an example of what your submission file should look like. The order of the rows does not matter, but the names of the "ID" must be correct.
Additional earth observation data.
Train contains the target. This is the dataset that you will use to train your model.
Additional earth observation data.
This is a starter notebook to help you make your first submission.
Test resembles Train.csv but without the target-related columns. This is the dataset on which you will apply your model to.
Additional earth observation data.
Train contains the target. This is the dataset that you will use to train your model.
Additional earth observation data.
This file describes the variables found in train and test.
Test resembles Train.csv but without the target-related columns. This is the dataset on which you will apply your model to.