Following the CDP Fall meeting on Saturday, October 13, 2018, a tutorial was given by Greg Mulholland and Josh Tappan of Citrine Informatics, a company based in Redwood City, CA, Pittsburgh, PA and Chicago, IL. Citrine Informatics has built an artificial intelligence platform for materials development which claims to empower users to develop new high-performance materials and chemicals faster than ever before. Citrine Informatics works with partners along the materials value chain to use state of the art data science techniques to identify areas of improvement and optimization in advanced materials discovery and production.
Topics covered included:
- why materials science has different challenges from most AI applications, how those challenges are being addressed today, and how they could be addressed in the future
- how AI can effectively fit into a materials development workflow as a powerful tool alongside traditional and high-throughput experimental and computational methods
- how to create, understand, and critically evaluate AI models on materials and chemicals data
- the role data plays in AI, and how to organize and structure data such that ML can be used most effectively
The workshop was intended to be an interactive introduction to a complicated topic. Greg and Josh made a substantial effort to directly address questions and topics of interest raised by the attending students and faculty from North Carolina State University and the Pennsylvania State University as well as CDP industry members.