Colorado School of Mines Partnership
University-level materials informatics curriculum content
The Mines Initiative for Data-Driven Materials Innovation (MIDDMI) is a collaboration between Colorado School of Mines and Citrine Informatics dedicated to educating Mines students on the fundamentals of materials informatics. During the spring semester of 2018, MIDDMI Fellows worked on materials informatics research problems, while implementing best practices for data management and applying machine learning models to materials research.
The Four Foundations of MIDDMI Research Projects
Research Data Best Practices
Applying machine learning to materials problems requires new approaches to data management. MIDDMI Fellows learned experimental and computational parameters so data could be used by other researchers. Current and subsequent datasets adhere to the FAIR data principles, which means data is Findable, Accessible, Interoperable, and Reusable. The FAIR Data Principles focus on enhancing data reusability and machine readability.
Clean, Structured Datasets
Building from the foundation of proper data capture and storage, Fellows learned how to structure data for input to machine learning models.
Machine Learning Models
Once research data was properly structured for machine learning, Fellows learned how to interpret and build machine learning models. These models were capable of guiding subsequent research, design, and discovery efforts.
MIDDMI Fellows learned about a data-driven approach to materials design called sequential learning. Sequential learning is an AI workflow with demonstrated success in reducing the number of experiments required to develop new high-performing materials. Sequential learning uses AI models to suggest which experiments to perform next to efficiently explore a design space.
Contact us to learn more about this materials informatics-focused curriculum.