Please use this identifier to cite or link to this item: http://hdl.handle.net/1822/18186

TitleApplication of data mining techniques to the safety evaluation of slopes
Author(s)Martins, Francisco F.
Miranda, Tiago F. S.
KeywordsData mining
Slope stability
Factor of safety
Probability of failure
Issue dateSep-2010
PublisherIOS Press
Abstract(s)In the present paper, Data Mining techniques has been applied to evaluate the stability of slopes. For this propose, the R (www.r-project.org) software was used together with a user defined application developed at the University of Minho called RMiner. The factor of safety (FS) and probability of failure (PF) were computed for 365 homogeneous slopes using the software SLOPE/W varying geometric parameters (height, height of the water surface and slope angle) and geotechnical parameters (weight density, cohesion intercept and angle of shearing resistance). Heights between 10 m and 15 m and slope angles between 40º and 70º were considered. This data allowed building a database to be analysed using the Data Mining techniques. In this process several algorithms were used for the prediction of FS and PF, namely: multiple regression, regression trees, artificial neural networks, support vector machines and k-nearest neighbours. To evaluate the performance of each technique REC curves (Regression Error Characteristic) and several error measures were used. This application allowed developing reliable models to predict important safety parameters for slopes without having to carry out a classical limit equilibrium calculation. They also allow performing quick parametric studies for the early stages of slope design. To predict FS the support vector machines showed to have the best overall performance. In the case of PF the artificial neural network proved to be more reliable to predict this parameter. With this study it was also possible to conclude that cohesion intercept was the parameter with more influence on the assessed safety parameters.
TypeConference paper
URIhttp://hdl.handle.net/1822/18186
ISBN978-1-60750-617-1-84
DOI10.3233/978-1-60750-617-1-84
Peer-Reviewedyes
AccessRestricted access (UMinho)
Appears in Collections:C-TAC - Comunicações a Conferências Internacionais

Files in This Item:
File Description SizeFormat 
Martins&Miranda-Shangai2010.pdf
  Restricted access
1,23 MBAdobe PDFView/Open

Partilhe no FacebookPartilhe no TwitterPartilhe no DeliciousPartilhe no LinkedInPartilhe no DiggAdicionar ao Google BookmarksPartilhe no MySpacePartilhe no Orkut
Exporte no formato BibTex mendeley Exporte no formato Endnote Adicione ao seu ORCID