arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.15214cond-mat.mtrl-sci
收藏

物理编码提升深度学习材料性能预测中的分布外性能

Physical Encoding Improves OOD Performance in Deep Learning Materials Property Prediction

Nihang Fu, Sadman Sadeed Omee, Jianjun Hu

AI总结:

本研究提出用物理编码替代独热编码,提升深度学习模型在材料性能预测中的分布外泛化性能,尤其对小型数据集训练模型效果显著。

摘要:

深度学习(DL)模型已被广泛应用于材料性能预测,并取得了巨大成功,尤其是对于拥有大型数据集的性能预测。然而,此类模型的分布外(OOD)性能值得质疑,特别是当训练集不够大时。在此,我们表明,使用物理编码而非广泛使用的独热编码,可以通过提升模型的泛化性能显著改善其分布外性能,这一点对于使用小型数据集训练的模型尤为明显。我们基于组成和结构的深度学习模型在六个数据集上的基准测试结果,包括形成能、带隙、折射率和弹性能预测,证明了物理编码对于小型数据集训练的模型的分布外泛化的重要性。

英文摘要:

Deep learning (DL) models have been widely used in materials property prediction with great success, especially for properties with large datasets. However, the out-of-distribution (OOD) performances of such models are questionable, especially when the training set is not large enough. Here we showed that using physical encoding rather than the widely used one-hot encoding can significantly improve the OOD performance by increasing the models' generalization performance, which is especially true for models trained with small datasets. Our benchmark results of both composition- and structure-based deep learning models over six datasets including formation energy, band gap, refractive index, and elastic properties predictions demonstrated the importance of physical encoding to OOD generalization for models trained on small datasets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑