| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Materials Science for Global Development -- Health, Energy, and Environment: An SMD Symposium in Honor of Wole Soboyejo
|
| Presentation Title |
Harnessing the Power of Machine Learning to Solve Global Problems |
| Author(s) |
Stephen Price, Winston Soboyejo, Rodica Neamtu |
| On-Site Speaker (Planned) |
Rodica Neamtu |
| Abstract Scope |
Recent advancements in technology and the growth of machine learning (ML) led to an abundance of data available for modeling. However, not all data is suitable for ML. This is particularly true for datasets collected in materials science labs and medical settings, often containing imbalances and missing values, due to specific diagnostic and experimental strategies. Such datasets lead to biased, skewed, or inaccurate ML models.
We address these challenges by introducing a novel framework to curate real-world datasets and make them suitable for ML, complemented by a ML-driven experimental design methodology capable of creating generalizable models that can be disseminated for rapid development of sustainable materials. Implemented on datasets familiar to the research community as well as datasets collected in our lab, our framework yields promising results and opens the door for expanding the use of ML to solve impactful problems such as climate change mitigation and affordable medical diagnostics. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Computational Materials Science & Engineering, |