Explosives are applied as a mining tools where mechanical processes are not applicable or economically feasible. The proximity and coexistence of these blasting processes with housing units present a problem in the operational context. Part of the non-absorbed energy from the rock blasting causes vibration and noise that propagated through the ground or the atmosphere, which can cause distress to the people living or working nearby. A methodology was proposed to predict the physical parameters resulting from the effects of ground vibration (PPV) and air overpressure (AOp) in safe blasting operations in a Gneiss open-pit mine in the South of Brazil. An engineering seismograph monitored shock waves for 31 detonations over three years. The seismograph reports obtained were evaluated using SPSS Statistics IBM 25 and Microsoft Excel 2016, to develop the multiple linear regression (MLR) models, and GeneXproTools 5.0 to build the shock wave prediction model using Gene Expression Programing (GEP). Based on the computational analysis was possible to create two methods of analysis and two computational models to predict shock waves: MLR and GEP. Through the performance analysis, the model obtained from the GEP proved to be more effective and assertive for predicting PPV and AOp.