Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Seismic Liquefaction Analysis of MCDM Weighted SPT Data Using Support Vector Machine Classification
Seismic soil liquefaction severely damages infrastructure and buildings. Identification of the key factors influencing seismic liquefaction is crucial since it will improve the prediction of evaluation models. In this article, multi-criteria decision making (MCDM) methods based on analytic hierarchy process (AHP) combined with other objective weighting methods such as entropy, criteria importance through inter-criteria correlation (CRITIC), standard deviation and method based on the removal effects of criteria (MEREC) were used to determine important parameters governing the seismic soil liquefaction such as earthquake magnitude (Mw), peak ground acceleration, normalised SPT blow count (N1,60), age of the deposit, effective over burden pressure (σv'), epicentral distance (r), relative density (Dr), D50, fines content (FC), clay content, plasticity index, ratio of water content to liquid limit, permeability coefficient (k). Further, the comparison analysis was made between four weighting methods in determining the important parameters governing seismic soil liquefaction. It was found that, epicentral distance (r), D50, fines content (FC) and permeability coefficient (k) have greater impact on the soil liquefaction. Recently developed MEREC method was found to be giving reliable outcomes in comparison to the other existing objective weighting methods. To validate the effect of parameter weights in determining the liquefaction susceptibility of soil, support vector machine (SVM) analysis is carried out on the standard penetration test (SPT) data set with and without considering the weights obtained from all the considered objective weighting methods. Performance of SVM models with Gaussian kernel has given better results with weights than without weights.
Seismic Liquefaction Analysis of MCDM Weighted SPT Data Using Support Vector Machine Classification
Seismic soil liquefaction severely damages infrastructure and buildings. Identification of the key factors influencing seismic liquefaction is crucial since it will improve the prediction of evaluation models. In this article, multi-criteria decision making (MCDM) methods based on analytic hierarchy process (AHP) combined with other objective weighting methods such as entropy, criteria importance through inter-criteria correlation (CRITIC), standard deviation and method based on the removal effects of criteria (MEREC) were used to determine important parameters governing the seismic soil liquefaction such as earthquake magnitude (Mw), peak ground acceleration, normalised SPT blow count (N1,60), age of the deposit, effective over burden pressure (σv'), epicentral distance (r), relative density (Dr), D50, fines content (FC), clay content, plasticity index, ratio of water content to liquid limit, permeability coefficient (k). Further, the comparison analysis was made between four weighting methods in determining the important parameters governing seismic soil liquefaction. It was found that, epicentral distance (r), D50, fines content (FC) and permeability coefficient (k) have greater impact on the soil liquefaction. Recently developed MEREC method was found to be giving reliable outcomes in comparison to the other existing objective weighting methods. To validate the effect of parameter weights in determining the liquefaction susceptibility of soil, support vector machine (SVM) analysis is carried out on the standard penetration test (SPT) data set with and without considering the weights obtained from all the considered objective weighting methods. Performance of SVM models with Gaussian kernel has given better results with weights than without weights.
Seismic Liquefaction Analysis of MCDM Weighted SPT Data Using Support Vector Machine Classification
Iran J Sci Technol Trans Civ Eng
Alla, Vamsi (Autor:in) / Sahoo, Upendra Kumar (Autor:in) / Behera, Rabi Narayan (Autor:in)
01.08.2024
11 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Seismic Liquefaction Analysis of MCDM Weighted SPT Data Using Support Vector Machine Classification
| Springer Verlag | 2024
Seismic liquefaction potential assessment by using Relevance Vector Machine
| Springer Verlag | 2007
Seismic liquefaction potential assessment by using Relevance Vector Machine
| Online Contents | 2007
Support vector machines-based modelling of seismic liquefaction potential
| British Library Online Contents | 2006
Seismic liquefaction potential assessed by support vector machines approaches
| Online Contents | 2015