20.05.2020 Publication on the methodology of machine learning in spatial prediction
Importance of the selection of spatial predictor variables in machine learning applications - from data reproduction to spatial prediction.
Hanna Meyer from the Nature 4.0 team, together with other authors, shows the importance of spatial cross-validation for reliable error estimates of predictions and how models can be significantly improved. Application examples are environmental variables in the Marburg Open Forest.
Click here for the publication: https://doi.org/10.1016/j.ecolmodel.2019.108815hh