Prediction of Slope Stability in Open-pit Mine Based on GA-LSSVR

Aiming at the small samples, nonlinear characteristics of slope stability problems in open-pit. Taking advantages of global search ability of genetic algorithm, a method of least squares support vector regression parameters optimized based on genetic algorithm is proposed, and a prediction model of slope stability in open-pit mine base on genetic least squares support vector regression (GA - LSSVR) is established. Genetic algorithm is applied to optimize the LSSVR，which can improve the accuracy and speed of the prediction. Experimental results demonstrate that the accuracy of GA-LSSVR is higher compared to the BP neural network and LSSVR model. The prediction method of slope stability in open-pit mine based on GA-LSSVR is more effective.