As part of the EMC Unified Convection (UC) parameterization that unifies the schemes (the Simplified Arakawa Schubert (SAS) and Grell-Freitas (GF)) using features from both schemes and optimizing the resulting convective parameterization for performance on regional and global scales. This unified scheme will use an ensemble of closures to determine the strength and location of convection. Deficiencies in optimizing the choice of the closures in different scales and changing environmental conditions, used in deep convection parameterizations in General Circulation Models (GCMs), have critical impacts on climate simulations. Currently, in the GF parametrization, a uniform average is taken over all closures. However, some closures may produce more accurate output in particular environmental or simulation conditions. By combining Machine Learning (ML) and satellite/global model convection datasets we can generate an optimal weighting vector based on location and environmental conditions. This vector will be used to weight the averaging of closure results for a given location, ensuring that the ones best suited to a particular environment, forecast time, or rainfall amount are most represented in the ensemble mean.