The original version of this Article contained an error in CODE AVAILABILITY,in which description of URL for the hyperlink is incorrect.In the corrected version,“https://github.com/XXXX/”is replaced by“https://gith...The original version of this Article contained an error in CODE AVAILABILITY,in which description of URL for the hyperlink is incorrect.In the corrected version,“https://github.com/XXXX/”is replaced by“https://github.com/nttcslab/floor-padding-BO.”This has been corrected in the PDF and HTML versions of the Article.展开更多
A crucial problem in achieving innovative high-throughput materials growth with machine learning,such as Bayesian optimization(BO),and automation techniques has been a lack of an appropriate way to handle missing data...A crucial problem in achieving innovative high-throughput materials growth with machine learning,such as Bayesian optimization(BO),and automation techniques has been a lack of an appropriate way to handle missing data due to experimental failures.Here,we propose a BO algorithm that complements the missing data in optimizing materials growth parameters.The proposed method provides a flexible optimization algorithm that searches a wide multi-dimensional parameter space.We demonstrate the effectiveness of the method with simulated data as well as in its implementation for actual materials growth,namely machine-learning-assisted molecular beam epitaxy(ML-MBE)of SrRuO_(3),which is widely used as a metallic electrode in oxide electronics.Through the exploitation and exploration in a wide three-dimensional parameter space,while complementing the missing data,we attained tensile-strained SrRuO3 film with a high residual resistivity ratio of 80.1,the highest among tensile-strained SrRuO3 films ever reported,in only 35 MBE growth runs.展开更多
文摘The original version of this Article contained an error in CODE AVAILABILITY,in which description of URL for the hyperlink is incorrect.In the corrected version,“https://github.com/XXXX/”is replaced by“https://github.com/nttcslab/floor-padding-BO.”This has been corrected in the PDF and HTML versions of the Article.
文摘A crucial problem in achieving innovative high-throughput materials growth with machine learning,such as Bayesian optimization(BO),and automation techniques has been a lack of an appropriate way to handle missing data due to experimental failures.Here,we propose a BO algorithm that complements the missing data in optimizing materials growth parameters.The proposed method provides a flexible optimization algorithm that searches a wide multi-dimensional parameter space.We demonstrate the effectiveness of the method with simulated data as well as in its implementation for actual materials growth,namely machine-learning-assisted molecular beam epitaxy(ML-MBE)of SrRuO_(3),which is widely used as a metallic electrode in oxide electronics.Through the exploitation and exploration in a wide three-dimensional parameter space,while complementing the missing data,we attained tensile-strained SrRuO3 film with a high residual resistivity ratio of 80.1,the highest among tensile-strained SrRuO3 films ever reported,in only 35 MBE growth runs.
