Evaluation of Deep Learning against Conventional Limit Equilibrium Methods for Slope Stability Analysis

This paper presents a comparison study between methods of deep learning as new category slope stability analysis, built upon the recent advances in artificial intelligence and conventional limit equilibrium analysis methods. For this purpose, computer code was developed to calculate factor safety (FS) using four methods: Bishop’s simplified method, Fellenius Janbu’s corrected method. The verified against Slide2 RocScience. Subsequently, average FS values were used approximate “true” slopes for labeling images learning. Using code, comprehensive dataset with wide ranges geometries soil properties created. label implementing two models: multiclass classification regression model. After training, models predict an independent set images. Finally, performance compared that found can reach accuracies high 99.71% while improving computational efficiency by more than 18 times