Current projections indicate the volume of NASA Earth science data will increase from 4 to 50 PB per year over the next several years.  Thus, unique, automated approaches to process these datasets is important for the application of detailed scientific and climatological studies to promote further scientific understanding.  Accuracy of these applications is further dependent on high quality phenomena specific datasets developed by subject matter experts for training, testing, and validation of machine learning (ML) algorithms. The purpose of this session it to provide an interdisciplinary forum to showcase existing methods and recent advancements of ML applications on big data repositories.  This session seeks contributions that leverage ML through both algorithm development and the development of tools to support ML database generation that promote the searchability and usability of large Earth science datasets.