In this study, using ground vibration data in the Sarcheshmeh Copper Mine, model using support vector regression intelligent method (SVR) and the fuzzy C-Means clustering (FCM) has been developed. To evaluate the performance of intelligent model, this model with experimental model United States Bureau of Mines (USBM) were compared. In addition, to demonstrate the power of clustering to improve results perdiction, models without the use of clustering techniques namely using the whole data have been developed. Therefore, In this study, four different models to predict the PPV was developed based on data Sarcheshmeh Copper Mine and their performance using statistical indicators RMSE, VAF and R 2 was examined. In all models, after the model developed by the training data with test data that form 15% the whole data, models have been examined. In the first model, six parameters including the burden, spacing, length of stemmeing, the number of holes in each delay, the amount of charge in each delay and the distance from the blasting using FCM was divided into three clusters and then taking into account two parameters the amount of charge in each delay and the distance from the blasting(Q and D) using SVR for each cluster model was developed to predict the PPV. Briefly in the first model, simultaneous clustering and SVR methods were used that concisely this model is called FCM-SVR. The second model is same as the first model with this difference that the empirical equation USBM has been used instead of SVR and because simultaneously clustering and USBM equation is used, this model is called FCM-USBM. In the third and fourth model, two parameters Q and D, without use of clustering (all data) as the inputs SVR and USBM methods has been used. In other words,in third and fourth model training data including total training of all three clusters and test data including total test data of all three clusters. Values of RMSE, VAF and R 2 , 1.8, 85.25 and 0.8536 for FCM-SVR model, respectively; 3.55, 49.06 and 0.4947 for FCM-USBM model respectively; 2.55, 70.75 and 0.7082 for SVR model respectively; and 4.05, 34.93 and 0.3541 for USBM model respectively was obtained. This results indicate that firstly SVR model is better than USBM model and secondly model FCM-SVR respect to SVR model is more accurate and generally, it can be seen that the clustering in accurate predicting ground vibration caused by blasting has an effective role