An advanced magnetic hysteresis modeling, exploiting the Preisach theory and the neural networks, is applied and discussed for the simulation of the magnetization processes of magnetic components made by laser powder bed fusion. Silicon iron samples with different percentage silicon content are used for the evaluation of the accuracy and reliability of the proposed approach. Measurements of the hysteresis loops and energy losses are compared with the computed results for different magnetization amplitudes and frequency rates. This approach is presented and evaluated here to provide a method for the accurate representation of the magnetization processes and the prediction of energy losses involved in additively manufactured ferromagnetic cores, in order to optimize their energy efficiency along different applications.