Modelling of shallow landslides with machine learning algorithms

Liu, Zhongqiang; Gilbert, Graham; Cepeda, José; Lysdahl, Asgeir Olaf Kydland; Piciullo, Luca; Hefre, Heidi; Lacasse, Suzanne

doi:10.1016/j.gsf.2020.04.014

Cited by 140 publications

(47 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A classification and regression procedure is established through a random group of variables selected at each tree node. To ensure reliable predictions, at least two conditions should be verified: the selected variables should have some predictive ability; the different decision tree models need to be uncorrelated [21].…”

Section: Random Forest (Rf)mentioning

confidence: 99%

Leak-Off Pressure Using Weakly Correlated Geospatial Information and Machine Learning Algorithms

et al. 2021

Self Cite

View full text Add to dashboard Cite

Leak-off pressure (LOP) is a key parameter to determine the allowable weight of drilling mud in a well and the in situ horizontal stress. The LOP test is run in situ and is frequently used by the petroleum industry. If the well pressure exceeds the LOP, wellbore instability may occur, with hydraulic fracturing and large mud losses in the formation. A reliable prediction of LOP is required to ensure safe and economical drilling operations. The prediction of LOP is challenging because it is affected by the usually complex earlier geological loading history, and the values of LOP and their measurements can vary significantly geospatially. This paper investigates the ability of machine learning algorithms to predict leak-off pressure on the basis of geospatial information of LOP measurements. About 3000 LOP test data were collected from 1800 exploration wells offshore Norway. Three machine learning algorithms (the deep neural network (DNN), random forest (RF), and support vector machine (SVM) algorithms) optimized by three hyperparameter search methods (the grid search, randomized search and Bayesian search) were compared with multivariate regression analysis. The Bayesian search algorithm needed fewer iterations than the grid search algorithms to find an optimal combination of hyperparameters. The three machine learning algorithms showed better performance than the multivariate linear regression when the features of the geospatial inputs were properly scaled. The RF algorithm gave the most promising results regardless of data scaling. If the data were not scaled, the DNN and SVM algorithms, even with optimized parameters, did not provide significantly improved test scores compared to the multivariate regression analysis. The analyses also showed that when the number of data points in a geographical setting is much smaller than that of other geographical areas, the prediction accuracy reduces significantly.

show abstract

Section: Random Forest (Rf)mentioning

confidence: 99%

Leak-Off Pressure Using Weakly Correlated Geospatial Information and Machine Learning Algorithms

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…In this paper, three different prediction models was developed for each of the three landslides. This section only briefly describes the machine learning algorithms used, as more details are available in (Cho et al, 2014;Yang et al, 2019b;Liu et al, 2020).…”

Section: Machine Learning Algorithms Implementedmentioning

confidence: 99%

“…The classification and regression procedure are established through a random group of variables selected at each tree node (Breiman, 2001). To ensure reliable predictions, at least two conditions should be verified: the selected variables should have some predictive power ability; the different decision tree models need to be uncorrelated (Liu et al, 2020). Detailed statistical explanation on RF algorithm is given in (Breiman, 2001).…”

Section: Random Forest Algorithmmentioning

confidence: 99%

“…Several techniques can be used to quantify the predictive capacity of influence factors including the information gain ratio (Tien Bui et al, 2016), the least support vector machine (Pham et al, 2018), and the Gini information gain (Quinlan, 1993). The Gini information gain method, using the RF approach (Liu et al, 2020;Zhang et al, 2020), was adopted in this study to assess the relative importance of each key factor. Information gain is used to determine which feature gives the maximum information for predicted results.…”

Section: Importance Of Key Factors Influencing Periodic Displacementmentioning

confidence: 99%

See 1 more Smart Citation

Algorithms for intelligent prediction of landslide displacements

Liu

Guo

Lacasse

et al. 2020

J. Zhejiang Univ. Sci. A

Self Cite

View full text Add to dashboard Cite

Landslides represent major threats to life and property in many areas of the world, such as the landslides in the Three Gorges Dam area in mainland China. To better prepare for landslides in this area, we explored how several machine learning algorithms (long short term memory (LSTM), random forest (RF), and gated recurrent unit (GRU)) might predict ground displacements under three types of landslides, each with distinct step-wise displacement characteristics. Landslide displacements are described with trend and periodic analyses and the predictions with each algorithm, validated with observations from the Three Gorges Dam reservoir over a one-year period. Results demonstrated that deep machine learning algorithms can be valuable tools for predicting landslide displacements, with the LSTM and GRU algorithms providing the most encouraging results. We recommend using these algorithms to predict landslide displacement of step-wise type landslides in the Three Gorges Dam area. Predictive models with similar reliability should gradually become a component when implementing early warning systems to reduce landslide risk.

show abstract

“…ere is a need for reevaluation of single ML models to prove their ability to produce good landslide susceptibility maps for their application in landslide-prone areas as they are simple to use in comparison to ensemble and hybrid models. Literature survey indicates that some of the single ML algorithms can compete with complex ensemble and hybrid models and can be used for the prediction of landslide susceptibility [17][18][19]. erefore, in the present study, we have used single linear discriminant analysis (LDA), logistic regression (LR), and radial basis function network (RBFN) methods for the landslide susceptibility mapping at Pithoragarh district of Uttarakhand state, India.…”

Section: Introductionmentioning

confidence: 99%

Landslide Susceptibility Mapping Using Single Machine Learning Models: A Case Study from Pithoragarh District, India

Trinh

Dam

Al‐Ansari

et al. 2021

Advances in Civil Engineering

View full text Add to dashboard Cite

Landslides are one of the most devastating natural hazards causing huge loss of life and damage to properties and infrastructures and adversely affecting the socioeconomy of the country. Landslides occur in hilly and mountainous areas all over the world. Single, ensemble, and hybrid machine learning (ML) models have been used in landslide studies for better landslide susceptibility mapping and risk management. In the present study, we have used three single ML models, namely, linear discriminant analysis (LDA), logistic regression (LR), and radial basis function network (RBFN), for landslide susceptibility mapping at Pithoragarh district, as these models are easy to apply and so far they have not been used for landslide study in this area. The main objective of this study is to evaluate the performance of these single models for correctly identifying landslide susceptible zones for their further application in other areas. For this, ten important landslide affecting factors, namely, slope, aspect, curvature, elevation, land cover, lithology, geomorphology, distance to rivers, distance to roads, and overburden depth based on the local geoenvironmental conditions, were considered for the modeling. Landslide inventory of past 398 landslide events was used in the development of models. The data of past landslide events (locations) was randomly divided into a 70/30 ratio for training (70%) and validation (30%) of the models. Standard statistical measures, namely, accuracy (ACC), specificity (SPF), sensitivity (SST), positive predictive value (PPV), negative predictive value (NPV), Kappa, root mean square error (RMSE), and area under the receiver operating characteristic curve (AUC), were used to evaluate the performance of the models. Results indicated that the performance of all the models is very good (AUC > 0.90) and that of the LR model is the best (AUC = 0.926). Therefore, these single ML models can be used for the development of accurate landslide susceptibility maps. Our study demonstrated that the single models which are easy to use and can compete with the complex ensemble/hybrid models can be applied for landslide susceptibility mapping in landslide-prone areas.

show abstract

Modelling of shallow landslides with machine learning algorithms

Cited by 140 publications

References 39 publications

Leak-Off Pressure Using Weakly Correlated Geospatial Information and Machine Learning Algorithms

Leak-Off Pressure Using Weakly Correlated Geospatial Information and Machine Learning Algorithms

Algorithms for intelligent prediction of landslide displacements

Landslide Susceptibility Mapping Using Single Machine Learning Models: A Case Study from Pithoragarh District, India

Contact Info

Product

Resources

About