| About this Abstract |
| Meeting |
MS&T21: Materials Science & Technology
|
| Symposium
|
Computation Assisted Materials Development for Improved Corrosion Resistance
|
| Presentation Title |
Understanding and Reducing Bias in Machine Learning to Enhance Its Predictive and Extrapolative Capabilities: Application to the Oxidation Kinetics and Spallation Behavior of High-temperature NiCr-based Alloys |
| Author(s) |
Marie Romedenne, Rishi Pillai, Jian Peng, Bruce Pint, Allen Haynes, Govindarajan Muralidharan, Dongwon Shin |
| On-Site Speaker (Planned) |
Marie Romedenne |
| Abstract Scope |
The development of new materials used in extreme environments needs a more profound understanding of the degradation of alloys in high-temperature oxidation environments. Combining modeling and experimental approaches such as machine learning (ML) with sufficient experimental data can accelerate the development of new materials while limiting their cost. In the current work, the role of the data distribution in the experimental dataset (data analytics), alloy composition, and chosen oxidation models (a simple parabolic law and a statistical cyclic-oxidation model) on the performance of ML models was evaluated. Potential strategies to improve the predictions and enhance the extrapolative capability of the previously trained model will be investigated. This research was sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, and the U. S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Vehicle Technologies Office, Propulsion Materials Program. |