National scale classification of landslide types by a data-driven approach and artificial neural networks

Classification of landslide type is important in risk management, yet it is often missing in large inventories. Here we present a novel data-driven method that uses morphometric and geospatial input parameters to classify landslides type at a national scale in Italy by means of a shallow Artificial Neural Network. The overall True Positive Rate is 0.76 for a five-class classification of over 275000 landslides. The performances in the entire national territory are very good, with F-score higher than 0.9 in large areas. The method can be applied to any polygonal inventory, as those produced by automatic mapping from Earth Observation imagery.