Towards overcoming data scarcity in materials science: unifying models and datasets with a mixture of experts framework
Abstract While machine learning has emerged in recent years as a useful tool for the rapid prediction of materials properties, generating sufficient data to reliably train models without overfitting is often impractical. Towards overcoming this limitation, we present a general framework for leveragi...
| المؤلفون الرئيسيون: | , , |
|---|---|
| التنسيق: | مقال |
| اللغة: | English |
| منشور في: |
Nature Portfolio
2022-11-01
|
| سلاسل: | npj Computational Materials |
| الوصول للمادة أونلاين: | https://doi.org/10.1038/s41524-022-00929-x |