Evaluation of Machine Learning Approaches to Estimate Aerosol Mixing State Metrics in Atmospheric Models