Using Fuzzy Neural Networks to Model Landslide Susceptibility at the Shihmen Reservoir Catchment in Taiwan

Machine learning algorithms are commonly employed in landslide susceptibility assessments. Recently, that utilize artificial intelligence have come into prominence. This study attempts to adapt the most fundamental framework of deep and introduces fuzzy theory concepts analyze while updating network parameters with trial-and-error methods. The final analysis results will compare those logistic regression (LR). In order assess ability model identify landslides a more objective way, two typhoon events were used as training event validation event, respectively. show area under curve (AUC) neural (FNN) for is 0.915, but AUC drops 0.746. Although FNN better than LR, they did not differ much from LR predicting future events. reason this difference between distributions too large, making biased its identification. Overall, still recommended method analyzing potential can be reference LR.