Since most of copper deposits in Iran are porphyry copper deposit,optimal exploration of porphyry copper deposits of Iran has great importance. In the explorationthese types of deposits, geophysical and geological perceptions are powerful techniques and provide useful information for researchers to determining thelocation of mineralization, separation high gradezones from background and determination optimal location of exploratoryboreholes. Due to thehigh complexity of earth structure and gang mineralization, interpreting and analyzing of this raw data is very difficult work that sometimes this complexity is confusing object for researcher. We propose automatic classification as a way to tackle this problem. This problem can be regarded as a classification problem that goal is discrete high grade zone from low grade zone (background). In the last decades researchers try to use the machine learning methods instead of classical methods that can reduce human error in the classification problems.Besides, automatic procedures are based on formalized data processing schemes, which render them reproducible and can be used in other similar situation. Thus, using multivariate statistical analysis and supervised classification methods can significantly reduce the risk level of decision making for the researchers. With the help of geophysical data including induce polarization, resistivity, magneticmeasurement and geologicaldataincludinglithologyandalteration, in this thesisforseparationhigh gradezones fromlow grade zones (waste) inporphyry copperdeposits, the three supervised classification methods include Support Vector Machine (SVM), Linear Discriminate Analysis (LDA) and Quadratic Discriminate Analysis (QDA) were used. Supervised classification methods have three steps, in first step the model have been trained and learned each example data belongs to which group with use of training data, afterwards in second step the model has been cross validate and in final step the model has been test by test data. Linear Discriminate Analysis and Quadratic Discriminate Analysis are two classical methods that have been used in a lot of variety classification studies in three last decades and expose good performance in majority of them. The goal of discriminant analysis is to obtain rules that describe the separation between groups of observations.Support Vector Machines are learning systems that use a hypothesis space of linear functions in a high dimensional feature space, trained with a learning algorithm from optimization theory that implements a learning bias derived from statistical learning theory. This learning strategy introduced by Vapnik and co-workers is a principled and very powerful method that in the few years since its introduction has already outperformed most other systems in a wide variety of applications and has good generalization ability compared to other methods. Informationanddataof twoAliabad and Darehzereshk porphyry copper deposit have beenanalyzed inthis thesis. In Aliabad deposit SVM method withaccuracyequal to80%incomparison QDA method with 69% accuracy and LDA method with67%accuracy, providedBetterperformanceinclassificationof data. In Darehzereshk depositSVM method with an accuracy of 90% in comparison QDA and LDA methods with 81% and 82% accuracy has beenmore successful in classification of data.