Machine Learning for Materials Genome and ICME Data
Managing the unprecedented growth in data-enabled science and engineering is a challenge for the modeling and simulations community in academia due to the tremendous diversity of data and lack of unifying principles in software development. In the proposed effort, we concentrate on the needs of rapidly growing materials research communities which also cuts across different disciplines. A new era of efficient supercomputing need an integrative platform that can augment learning process, and provide an incentive to the users to interact and adopt best practices for emerging supercomputing architectures. A campus-wide collaboration is launched with engineering, computing and data science faculties to build an efficient platform with machine learning capabilities. This issue in in particular prominent in materials genome research where analyzing existing data sets for obtaining scientific insights is a grand challenge and the volume of data is increasing at a rapid rate.
