Abstarct: Intelligent methods and pattern recognition are the most common methods that have been used for data analysis in science and engineering fields. These methods have been used for anomaly of background separation in exploration of geochemistry. In the current research, clustering and classification algorithms were used together as a pattern recognition method using surface and depth information to estimate concentration of the elements in the missing points and different depth horizons. The data were clustered and the concentrations of the elements were estimated using the neural network and clustering methods. For this purpose, the number of optimum clusters was identified baxsed on validation indices. The results showed that the combination of FCM and clustering algorithms increased the accuracy of the estimation from 75 up to 88 percents. The geochemical maps at different depth levels were drawn. Front, tail and near ore haloes were identified. The behavior of elements showed that the arsenic as front halo and vanadium as tail halo reacted. baxsed on the classification of copper concentration changes by fractal-grade-area method and determining copper classes with classification algorithms, the model of the deposit was depicted in depth. The maps showed that the ore deposit was eroded in the southwestern part of the area, and there was another mineralized zone in depth at the north-central part of the region. Among the different used classification methods, the highest accuracy was related to the neural network algorithm with precision classification of 90.6%. The geophysical maps and sections together with drilling data were used for validation of the results. These data confirmed the geochemical results.