Establishment of a deformation forecasting model for a step-like landslide based on decision tree C5.0 and two-step cluster algorithms: a case study in the Three Gorges Reservoir area, China
- 582 Downloads
This study presents a hybrid approach based on two-step cluster and decision tree C5.0 algorithms to establish a deformation forecasting model for a step-like landslide. The Zhujiadian landslide, a typical step-like landslide in the Three Gorges Reservoir area, was selected as a case study. Approximately, 6 years of historical records of landslide displacement, precipitation, and reservoir level were used to build the forecasting model. The forecasting model consisted of seven comprehensive rules governing hydrologic parameters and their magnitudes and was developed to predict landslide deformation. This model was applied to rapidly forecast the likelihood of step-like landslide deformation resulting from rainfall and water level fluctuations in the Three Gorges Reservoir area. Given the satisfactory accuracy of the trained model, the presented approach can be used to establish forecasting models for step-like landslides and to facilitate rapid decision making.
KeywordsStep-like landslide Deformation forecasting Two-step cluster Decision tree C5.0 Three Gorges Reservoir area
Author Junwei Ma thanks the China Scholarship Council for providing a scholarship for the research described in this paper, which was conducted as a Visiting Research Scholar at Purdue University. This study was financially supported by the Key National Natural Science Foundation of China (41230637), the National Natural Science Foundation of China (41572279, 41272305, and 41102195), the China Postdoctoral Science Foundation (Grant Nos. 2012M521500 and 2014T70758), and the Hubei Provincial Natural Science Foundation of China (Grant No. 2014CFB901). All support is gratefully acknowledged.
- Cao Y, Yin KL, Alexander D, Zhou C (2016) Using an extreme learning machine to predict the displacement of step-like landslides in relation to controlling factors. Landslides 13:725–736. doi: 10.1007/s10346-015-0596-z
- Huang HF, Yi W, Lu SQ, Yi QL, Zhang GD (2014) Use of monitoring data to interpret active landslide movements and hydrological triggers in three gorges reservoir. J PERFORM CONSTR FAC C4014005. doi: 10.1061/(ASCE)CF.1943-5509.0000682
- Liu XY, Li P, Gao CH (2013) Symmetric extreme learning machine. Neural Comput Appl 22(3-4):551–558. doi: 10.1007/s00521-012-0859-8
- Qin B, Xia Y, Li F (2009) DTU: a decision tree for uncertain data. In: Theeramunkong T, Kijsirikul B, Cercone N, Ho T-B (eds) Advances in knowledge discovery and data mining: 13th Pacific-Asia conference, PAKDD 2009. LNCS, vol. 5476. Springer, Heidelberg, pp 4–15. doi: 10.1007/978-3-642-01307-2_4 CrossRefGoogle Scholar