2025-09-04, 15:30–17:00 (Europe/Amsterdam), HugoTECH
In this tutorial, an AutoML framework (EvalML) will be used for modeling livestock density at polygon level (areal regression) and produce spatial predictions at pixel level.
A compilation of subnational livestock census data will be used to retrieve the headcount information, while the grazing areas will be estimated based on GPW grassland class and extent. You will learn how to run benchmark, feature selection and hyper-parameter tuning for several modeling pipelines, as well as evaluate the model performance using robust spatial cross-validation. The final modeling output will be spatial layer with livestock density and headcount at 1-km for pilot area.
Other links:
https://docs.google.com/presentation/d/1qePs0puoVaVaVysztSlqLadb4_YIxQNf2JeuaF5GWlI
