The methodology for scaling the geometry and operating parameters of Hall Thrusters, as described in this paper, relies on supervised Machine Learning algorithms. This approach aims to determine the geometrical dimensions, propellant mass flow rate and discharge voltage, with thrust and specific impulse as input requirements. The magnetic field is also taken into account. Gradient Boosting Regression (GBR) is identified as the most appropriate algorithm for this purpose. The scaling process utilizes a specific database of 54 thrusters, incorporating measurements with xenon, krypton, and argon as propellants. An analytical method based on the GBR and optimization algorithms has been developed and validated to design a Hall thruster that meets space mission specifications. Additionally, the procedure is complemented by a more direct and intuitive graphical method.