Ore Waste Classification Of A Lead Zinc Deposit Using Support Vector Machine

The objective of this study is to classify ore and waste in a selected mine. The support vector machine, was used for the classification purpose, is applied to a zinc ore body in a skarn deposit located in the central region of the Peruvian Andes. The input parameters used for the model are spatial coordinates and the lithological information of the ore body. The input data was divided into training and testing data sets and the performance of the model was tested using the testing data set. The results show that more than 76% of the data could be properly classified by this method. The ore waste maps of the deposit was then developed using SVM model. From the result it was observed that a comparatively very few number of cells classified as ore. The ore is concentrated in small lodes at the middle of the volume of study, and not uniformly distributed throughout the study area. The lithological map of the deposit was also constructed as it was used as an input parameter for the SVM model. The indicator kriging was used for generation of lithological map of the ore body. Key word: Ore waste classification, Support vector machine, Indicator kriging, Confusion matrix.