| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Innovations in Energy Materials: Unveiling Future Possibilities of Computational Modelling and Atomically Controlled Experiments
|
| Presentation Title |
The Magic and Myths of Machine Learning in Materials Science |
| Author(s) |
Rika Kobayashi, Emily Kahl, Roger D Amos |
| On-Site Speaker (Planned) |
Rika Kobayashi |
| Abstract Scope |
Though the concept of Machine Learning (ML) has been around since 1959 it has only been relatively recently that it has started to pervade the applied sciences. Carried along by perceived successes in computer vision and large language models ML is being applied in an increasing number of projects in a growing number of application areas. Resulting publications make grand claims on the power of ML amongst some genuinely useful discoveries. In this talk I will give an overview of the various areas in which ML is being used in Materials Science. I will follow by focusing on two areas we have been concentrating recent efforts. The first will go over our extensive investigation of the applicability of Machine Learning Interatomic Potentials. The second part will introduce our work on training large language models for extracting useful data and information from the Materials Science corpus. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Computational Materials Science & Engineering, Modeling and Simulation |