Search

Based on the machine learning algorithms, prediction models for landslide displacement with step-like behavior in the reservoir area were established for landslides prevention and reduction; these models could predict a given test set very well. However, due to the length and the sequence of the training set in prediction models, the predictive ability of these prediction models could not be evaluated accurately if just validated with a given test set. To solve the problem, a hybrid reliability model was proposed. Complimentary ensemble empirical mode decomposition (CEEMD) algorithm was used to decompose the accumulated displacement into the trend displacement and the periodic displacement firstly. The Gauss function was used to predict the trend displacement, and the random forest (RF) and back propagation neural network (BPNN) algorithms were employed to predict the periodic displacement. Furthermore, a novel performance function for the reliability analysis of the displacement prediction model was derived to address failure probabilities in different cases. The Baijiabao landslide in the Three Gorges Reservoir Area was taken as an example for reliability analysis of the prediction model for landslide displacement with step-like behavior, and the predictive ability of the CEEMD-RF model and the CEEMD-BPNN model were compared. The results indicated that the CEEMD-RF model and CEEMD-BPNN model both can accurately predict the accumulated displacement of the given test set; the predictive value obtained with the CEEMD-RF model or the CEEMD-BPNN model showed uncertainties of the prediction model for landslide displacement and the predictive ability of the CEEMD-RF model was more reliable than the CEEMD-BPNN model under different cases. The failure probability proposed in the paper could evaluate the predictive ability of the model more accurately and comprehensively compared with the existing assessment indices. [ABSTRACT FROM AUTHOR]

Jurassic facility-sliding strata have been identified as a fundamental factor affecting the occurrence of rainfall-reservoir induced landslides in western Hubei Province, China Three Gorges Reservoir area. Regional landslide susceptibility mapping is the most effective method for landslide prediction and mitigation. To solve the current problem of identifying the true landslides and non-landslides, a novel hybrid model based on the two steps self-organizing mapping-random forest (two steps SOM-RF) algorithm is proposed. The identified high and very high susceptibility zones are located within the hydro-fluctuation belt and regions with low altitude based on the datasets before 2014. Deviation and variance of other ten datasets are generated to evaluate the reliability of the maps for the problem of unbalanced sample sizes. Two typical landslides occurred in 2017 have been found in the very high susceptibility zone, which emphasized the validity of susceptibility mapping. To verify the effectiveness of selecting true landslides and non-landslides based on two steps SOM model, recorded landslides and non-landslides randomly chosen from landslide-free areas are put into the single RF model for comparison. The receiver operating characteristic curve and Accuracy index are applied to compare the performance of the landslide susceptibility maps created based on two steps SOM-RF and the single RF model. The results demonstrate that the consideration of true landslides and non-landslides is effective in producing a more accurate landslide susceptibility map with superior prediction skill and higher reliability. [ABSTRACT FROM AUTHOR]