Multiphase particle in a cell simulation of fluidized beds for different Geldart group particles (A, B and D) were investigated. The minimum fluidization velocities predicted by the Barracuda Virtual Reactor (R) were compared and contrasted by theoretical predictions with the Wen-Yu model. The optimal drag force coefficients based on the Wen-Yu model for MP-PIC simulations was further evaluated for different particle sizes and densities for Group A catalysts. A neural network model was constructed for the optimal model constants as a function of the explanatory variables for group A catalysts. This work illustrates the importance of choosing the correct value for the drag force coefficients to obtain realistic simulation results for fluidized beds. This work clearly demonstrates that the Barracuda simulation results are very sensitive to the value of the Wen-Yu parameter. Using multiphase particle in cell simulations in conjunction with a neural network model can be directly used in further studies as a rapid tool to get the correct values for realistic drag force coefficients based on the Wen-Yu model.