CaRS is currently collaborating with the Department of Computer Science at NPS (Dr. Marko Orescanin) to apply various machine learning methods to remote sensing data. The two groups share some of their students, who are involved in development of Bayesian neural network design for classification and regression problems. Some examples of what our groups have been working on include 1) rain-type classification using passive microwave radiances, 2) prediction of microwave brightness temperature from geostationary radiances, and 3) prediction of radar reflectivity or rain rate using geostationary radiances. Recent work seeks to also actively predict the uncertainty for both classification and regression techniques. CaRS provides support for processing remote sensing data and ingesting it into neural networks.