| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Integrating Experimental Data into Dynamic Artificial Intelligence/Machine (AI/ML) Learning Workflows |
| Author(s) |
Elizabeth Ann Pogue, Ann Choi, Denise Yin, Michael Pekala, Nam Le, Alexander New, Eddie Gienger, Christian Sanjurjo-Rodriguez, Bianca Pilosino, Douglas Trigg, Anna Langham, Georgia Leigh, Sebastian Lech, Gregory Bassen, Elizabeth Hedrick, Brandon Wilfong, Steven Storck, Mitra Taheri, Tyrel M. McQueen, Christopher Stiles |
| On-Site Speaker (Planned) |
Elizabeth Ann Pogue |
| Abstract Scope |
Data is at the heart of AI and ML, serving both as the foundation for model training and as the benchmark for evaluating model accuracy. Good data costs money and requires time and effort to gather and curate. We discuss the challenges associated with and insights gained from developing material synthesis and characterization approaches aimed at high-throughput materials discovery. While a variety of datasets such as images, text, and small databases, are readily available for training diverse models now, the continuous advancements in AI and ML require the integration of new data sources and the merging of disparate databases and collections. Our goal is to produce high-quality, cost-effective experimental data that meets the current needs in fields such as superconductivity, multi-principal component magnets, and structural alloys, and to package these data collections in ways that enable future studies unimaginable today. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Entropy Alloys, Machine Learning, Magnetic Materials |