Research Article
Classification of Earthquake-Induced Damage for R/C Slab
Column Frames Using Multiclass SVM and Its Combination
with MLP Neural Network
Ali Kia and Serhan Sensoy
Department of Civil Engineering, Eastern Mediterranean University, Gazimagusa, Mersin 10, Turkey
Correspondence should be addressed to Ali Kia; ali kia cien@yahoo.com
Received 4 March 2014; Revised 3 June 2014; Accepted 5 June 2014; Published 23 July 2014
Academic Editor: Xuejun Xie
Copyright © 2014 A. Kia and S. Sensoy. Tis is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Nonlinear time history analysis (NTHA) is an important engineering method in order to evaluate the seismic vulnerability of
buildings under earthquake loads. However, it is time consuming and requires complex calculations and a high memory machine.
In this study, two networks were used for damage classifcation: multiclass support vector machine (M-SVM) and combination of
multilayer perceptron neural network with M-SVM (MM-SVM). In order to collect data, three frames of R/C slab column frame
buildings with wide beams in slab were considered. For NTHA, twenty diferent ground motion records were selected and scaled to
ten diferent levels of peak ground acceleration (PGA). Tus, 600 obtained data from the numerical simulations were applied to M-
SVM and MM-SVM in order to predict the global damage classifcation of samples based on park and Ang damage index. Amongst
the four diferent kernel tricks, the Gaussian function was determined as an efcient kernel trick using the maximum total accuracy
method of test data. By comparing the obtained results from M-SVM and MM-SVM, the total classifcation accuracy of MM-SVM
is more than M-SVM and it is accurate and reliable for global damage classifcation of R/C slab column frames. Furthermore, the
proposed combined model is able to classify the classes with low members.
1. Introduction and Background
Artifcial neural networks (ANNs) are one of the popu-
lar computational models applied widely throughout the
sciences. Also they are specially used in many felds of
civil engineering like materials strength prediction, thermo-
graphic inspection of electrical installations within buildings,
trafc management and transportation systems, forecast
water pressure in pipes, and so forth. Generally ANNs are
used to solve complex problems by considering efective
indices and establish a good relationship between the input
and output parameters. Moreover, these networks can be
applied in classifcation problems. Te frst classifcation
algorithm was presented by Fisher [1]. In this algorithm,
minimizing the classifcation error of train data was eval-
uated as an optimization criterion. Tis method has been
used in many classifcation algorithms, yet there are some
problems encountered mainly the generalization properties
of the classifers, which are not directly involved in the cost
function. Also for doing the training process, determining
the structure of the network was not easy. As an example,
to determine the optimum number of neurons in the hidden
layers of the multilayer perceptron (MLP) neural networks
or the number of Gaussian functions in radial basis function
(RBF) neural networks are difcult and time consuming.
Cortes and Vapnik [2] introduced a new learning statistical
theory which leads to presenting the support vector machines
(SVMs). Te signifcant features of these networks are their
ability to minimize the classifcation errors, maximize the
geometric margins between classes, design the classifers
with maximum generalization, and automatically determine
the architecture of network for classifers and modeling the
nonlinear separator functions using nonlinear cores.
In recent years, several diferent neural networks such as
SVM have been applied in diferent branches of civil engi-
neering. In a tunnel construction, an intelligent controlling
system was presented by Jun et al. [3]. Tis system needed
to recognize the geophysical parameters to fnd the optimum
Hindawi Publishing Corporation
Mathematical Problems in Engineering
Volume 2014, Article ID 734072, 14 pages
http://dx.doi.org/10.1155/2014/734072