One implementation of SVM in EC 6 compliant wall compressive strength classification

Abstract

In this paper, the possibilities of using the Support Vector Machine (SVM) for the classification of a complex set of input data that define the problem of calculating the compressive strength of a wall in construction are analyzed. The compressive strength of the wall is determined by applying the empirical expression given by Euro Code 6 (EC 6), by the European Commission for Standardization, which fully defines the process of designing masonry structures. The aim of the work is to analyze the classification using SVM in relation to different kernels and to assess the quality of the prediction based on that solution. Finding an appropriate methodology for classification and prediction, an input vector of data that does not have a precalculated compressive strength of the wall should be classified against a previously trained algorithm and as such clearly indicate to construction experts which set of input parameters gives the most favorable result in order to define the strength at wall pressure.