期刊
PHYSICAL REVIEW B
卷 105, 期 16, 页码 -出版社
AMER PHYSICAL SOC
DOI: 10.1103/PhysRevB.105.165141
关键词
-
资金
- European Union [824143]
- European Research Council (ERC) under the European Union [890414]
- Marie Curie Actions (MSCA) [890414] Funding Source: Marie Curie Actions (MSCA)
This passage discusses the characteristics of machine learning potentials built upon high-dimensional atom-density representations, indicating that the probability density induced by training points in the representation space plays a significant role in robust extrapolation and accurate prediction.
We show that, contrary to popular assumptions, predictions from machine learning potentials built upon highdimensional atom-density representations almost exclusively occur in regions of the representation space which lie outside the convex hull defined by the training set points. We then propose a perspective to rationalize the domain of robust extrapolation and accurate prediction of atomistic machine learning potentials in terms of the probability density induced by training points in the representation space.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据