Application of Machine Learning Techniques in Slope Failure Analysis

With furtherance in the mining industry, accidents due to slope failure are getting frequent in mining sites. Slope instability being a complex process, it seriously threatens the miner’s life and properties. The damage inflicted by slope failures in the recent past has pulled the attention of authorities toward implementing disaster risk reduction measures. This research work plays a dominant role in palliating the slope failure risk. The presented work demonstrates the potentiality of machine learning models in forecasting the stability of the slopes. We implemented the limit equilibrium method (LEM) in predicting the factor of safety (FOS) of the slip surfaces for the designed slope. With the application of machine learning (ML) models such as K nearest neighbours and Gaussian Naive Bayes, we further classified the slopes based on their degree of stability. The performance of ML models is examined and compared through quality metric parameters like accuracy and confusion matrix.