Técnicas para la predicción espacial de zonas susceptibles a deslizamientos

The implemented techniques for the prediction of landslide-prone areas have been effective at a certain degree. However, many approaches tend to face difficulties to determine non-linear landslides triggering factors, due to the absence of Spatio-temporal dependency structures that evaluate spatial effects as autocorrelation and heterogeneity when describing complex problems. Therefore, results understanding may not be precise and lead to a less reliability condition. The main objective of this article is to provide a solid document that offers both, a general and a detailed perspective about Spatial Prediction Techniques. Finally, we propose an innovative methodology that allows us to use automatic learning and spatial statistics to improve the predictive performance of landslide-prone areas.