A comparative study of decision tree and support vector machine methods for gold prospectivity mapping

Elements geochemical anomalies mapping is one of the main goals of geochemical investigations. Since the field studies is time-consuming and costly, the data processing, applying machine learning methods could be applied, which are less expensive, faster, and more accurate. In this study, two machine learning methods, including decision tree algorithm and support vector machine, were applied for gold prospectivity mapping. To reach this aim, silver, copper, lead and zinc values were used as input variables. The comparison of the analytical results obtained from the two mentioned methods confirms that the DT (decision tree) model has sufficient accuracy and better performance than the SVM (support vector machine) model for preparing gold prospectivity mapping in the studied area.