Prediction of Degree of Soil Contamination Based on Support Vector Machine and K-Nearest Neighbor Methods: A Case Study in Arak, Iran

The degree of soil contamination in an urban region can be changed by heavy metals. This might result in endangering safety of an urban region. This paper presents an approach to build a prediction model for the assessment of degree of contamination index, based upon heavy metals changes. The heavy metal concentration of Pb, Cu, Ni, Zn, As, Cr and Ni as input was used to build a prediction model for the assessment of degree of contamination. Two prediction models were implemented such as support vector regression (SVR) and k-nearest neighbor regression method (KNNR). A comparison was made between these two models and the results showed the superiority of the SVR model. Furthermore, a case study in Arak, Iran was conducted to illustrate the capability of the support vector machines (SVM) model.