Recent and on-going efforts have shown how machine learning (ML) techniques, incorporating more, and more disparate data than can be interpreted manually, can predict seafloor properties, with uncertainty, where they have not been measured directly. Our program seeks to improve our global prediction of fundamental seafloor properties affecting acoustic bottom interaction, load-bearing capacity, and seafloor stability. We are particularly interested in predicting fundamental properties such as porosity, fraction clay, gas and gas hydrate saturation, and stress state, for which known deterministic models can then be used to constrain estimates of many other quantities. Our research in ML also ties in with other deterministic seafloor modeling, particularly numerical simulation of fluid flux through porous media as a means of solute transport from seafloor sediments to the ocean and atmosphere system