Using the data set containing CFD results and mass properties from a heat exchanger model, train a surrogate model capable of predicting pressure drop, average flow velocity, core surface area, and mass given input lattice cell size in the X-direction and Y/Z-direction, and inlet flow velocity. Once your model is trained, use inverse design to specify the optimal lattice cell sizes and flow velocity to maximize the surface area while satisfying the specified constraints for pressure drop, mass, and average flow velocity. The value of a surrogate model is measured in both accuracy and speed, so you will also be evaluated on how long it takes for your model to generate a prediction given input parameters.