Prediction of Rainfall Induced Landslides Using Machine Learning Models along Highway-Bandipora to Gurez Road, J&K, India

The present study endeavours to explore the efficacy of machine learning models in landslide predictions triggered by extreme rainfall events along Highway from Bandipora to Gurez, J&K, India. Random Forest (RF) and Logistic Regression (LR) Models were employed on account of their technical utility and logical procedure to transform the decision making into a series of algorithm steps that eventually culminate in low variance and minimum bias. The crux of these models was to find the optimal parameters for targeted feature i.e. landslide prediction. Hyper-parameter tuning was used to optimize the performance of the respective models. These models were evaluated for accuracy using the receiver operating characteristics, area under the curve (ROC-AUC) and false negative rate (FNR). When antecedent rainfall data is incorporated in the prediction models, both (LR and RF) performed better with an AUC of 0.825 and 0.843 respectively. The results reveal a positive correlation between antecedent precipitation and landslide occurrence rather than between single-day landslide and rainfall events. In case of FNR, LR and RF improved to14.32% and 15.92% respectively, when antecedent rainfall was included in the analysis. Comparing the two models, LR model’s performance is well within the acceptable limits of FNR and therefore could be preferred for landslide prediction and early warning over RF. LR model’s incorrect prediction rate is 8.48% without including antecedent precipitation data and 5.84% including antecedent precipitation data. Our study calls for wider use of Machinery Learning Models for developing early warning systems of landslides.

Keywords: -Rainfall Triggered; Machine Learning Models; Area Under Curve; False Negative Rate.