Prediction of the disaster area of loess landslide based on least square support vector machine optimized by bat algorithm

The prediction of landslide disaster area has always been one of the difficulties in landslide research. The loess landslides in South Jingyang plateau were chosen to establish model of disaster area prediction, by electing height, volume, source area length and width of landslide as the influencing factors, which based on the bat algorithm to optimize calculation for least squares support vector machine in the regularization parameters (γ and σ₂). In the meantime, they are compared with mutiple linear regression model. The result shows that the model has better prediction accuracy and effect. It can be used as the basis for disaster prevention and reduction in the area.