Sahel - Cattle Theft Probability Prediction

Data and Resources [1]
Additional information
Time Period of the Dataset [?]
January 01, 2018-September 30, 2024 ... More
Modified [?]
8 October 2024
Dataset Added on HDX [?]
8 October 2024 Less
Expected Update Frequency
As needed
Location
Source
Acción contra el hambre - GIS4tech
Methodology

The model trains with the actual data of the variables to be predicted and uses their correlation with agroclimatic indicators, biomass, and violence to predict said variable where there is no actual data, at the level of Commune/Municipality for all of Senegal and Mauritania.

The accuracy of the model is the fraction of predictions that the model made correctly.

A study is carried out to determine which model provides the best performance in terms of accuracy. The best performing model was the Hist Gradient Boosting Classifier, a histogram-based Gradient Boosting Classification Tree used for ranking, classification and many other machine learning tasks.

The model has been trained with historical data to subsequently make predictions for each Commune/Municipality for each month.

Caveats / Comments
File Format
Visibility
Public
Export metadata for this dataset: JSON | CSV