Moving closer to experimental level materials property prediction using AI.

Moving closer to experimental level materials property prediction using AI.
复制标题

DOI:
10.1038/s41598-022-15816-0
复制
发表时间:
2022-07-13
期刊:
影响因子:
4.6
通讯作者:
Agrawal, Ankit
Agrawal, Ankit
中科院分区:
综合性期刊3区
文献类型:
--
作者:
Jha, Dipendra;Gupta, Vishu;Liao, Wei-keng;Choudhary, Alok;Agrawal, Ankit

文献摘要

参考文献

被引文献

相似文献

虽然实验和DFT计算一直是理解晶体材料的化学和物理性质的主要手段,但实验是昂贵的,DFT计算是耗时的,并且与实验有显著的差异。目前,基于DFT计算的预测建模为进一步的DFT计算和实验提供了一种快速筛选候选材料的方法;然而,这种模型继承了基于DFT的训练数据的巨大差异。在这里,我们展示了如何利用AI与DFT一起,通过专注于预测“给定其结构和组成的材料的形成能量”的关键材料科学任务,比DFT本身更准确地计算材料属性。在包含137个条目的实验保持测试集上,AI可以从材料结构和成分预测形成能,平均绝对误差(MAE)为0.064 eV/原子;将其与DFT计算进行比较,我们发现AI可以首次显著优于DFT计算。
While experiments and DFT-computations have been the primary means for understanding the chemical and physical properties of crystalline materials, experiments are expensive and DFT-computations are time-consuming and have significant discrepancies against experiments. Currently, predictive modeling based on DFT-computations have provided a rapid screening method for materials candidates for further DFT-computations and experiments; however, such models inherit the large discrepancies from the DFT-based training data. Here, we demonstrate how AI can be leveraged together with DFT to compute materials properties more accurately than DFT itself by focusing on the critical materials science task of predicting “formation energy of a material given its structure and composition”. On an experimental hold-out test set containing 137 entries, AI can predict formation energy from materials structure and composition with a mean absolute error (MAE) of 0.064 eV/atom; comparing this against DFT-computations, we find that AI can significantly outperform DFT computations for the same task (discrepancies of eV/atom) for the first time.
DOI: 10.1038/s41598-021-83193-1
发表时间: 2021-02-19
期刊: Scientific reports
影响因子: 4.6
作者:
Jha D;Gupta V;Ward L;Yang Z;Wolverton C;Foster I;Liao WK;Choudhary A;Agrawal A
通讯作者: Agrawal A
DOI: 10.1038/srep34256
发表时间: 2016-10-03
期刊: Scientific reports
影响因子: 4.6
作者:
de Jong M;Chen W;Notestine R;Persson K;Ceder G;Jain A;Asta M;Gamst A
通讯作者: Gamst A
DOI: 10.1063/1.4946894
发表时间: 2016-05-01
期刊: APL MATERIALS
影响因子: 6.1
作者:
Agrawal, Ankit;Choudhary, Alok
通讯作者: Choudhary, Alok
DOI: 10.1103/physrevlett.117.135502
发表时间: 2016-09-20
影响因子: 8.6
作者:
Faber, Felix A.;Lindmaa, Alexander;Armiento, Rickard
通讯作者: Armiento, Rickard
DOI: 10.1038/s41598-018-35934-y
发表时间: 2018-12-04
期刊: SCIENTIFIC REPORTS
影响因子: 4.6
作者:
Jha, Dipendra;Ward, Logan;Agrawal, Ankit
通讯作者: Agrawal, Ankit