[16a-M_B104-1~10] 6.6 Probe Microscopy

12:00 PM - 12:15 PM JST(3:00 AM - 3:15 AM UTC)Presentation by Applicant for JSAP Young Scientists Presentation Award

[16a-M_B104-9] Machine-Learning-Based Objective Analysis of STS Spectra

〇(DC)Haruto Seki¹, Mai Niida¹, Kenji Nawa², Chiharu Mitsumata³, Toyo Kazu Yamada^1,4 (1.Chiba Univ., 2.AIST, 3.Tsukuba Univ, 4.Chiba Univ Mol. Chiral. Res.)

Keywords:

Scanning Tunneling Microscopy,Scanning Tunneling Spectroscopy,Machine learning

In conventional STS analysis, spectral classification often relies on visual inspection, which limits objectivity and reproducibility. In this study, we propose an objective analysis method that enables feature extraction and quantitative classification of a large number of STS spectra using machine learning. In STS map, accurate classification is challenging because the spatial distribution of electronic states varies with energy. To overcome this issue, we perform classification using spectral information over the entire energy range, thereby achieving energy-independent mapping of electronic states.