[23p-B200-1~5] Innovation and application of measurement informatics

[23p-B200-3] Application of informatics for thin-film science -from formation to spectral analysis-

〇Naoka Nagamura^1,2,3 (1.NIMS, 2.JST PRESTO, 3.Tokyo Univ. of Science)

Keywords:imaging, spectroscopy, machine learning

Machine learning techniques are powerful tools for quantitative analysis of imaging outputs in experiemental measurements. In the case of thin-film science, the surface structure can be confirmed from RHEED pattern images during the film formation process, and the effects of defects, grain boundaries, and device structures for material characteristics can be evaluated by spectromicroscopy techniques. Now we have developed a highly efficient, accurate, and fast analysis method by using machine learning for these imaging analysis. In the presentation, we will talk about our recent results.