[16a-A36-1~5] Advanced Technology on the Cyber-Physical System for the Society 5.0 -Connecting with the physical spaces as the technologies of Materials, Devices, Processes, Circuits, and Applications-

10:15 AM - 10:45 AM JST(1:15 AM - 1:45 AM UTC)

[16a-A36-2] Development of Machine Learning Potentials Using Persistent Homology

〇Emi Minamitani¹ (1.SANKEN, Osaka Univ.)

Keywords:

Topological Data Analysis,Machine Learning Potential

In recent years, attempts have been made to link the prediction of material properties by using information obtained from persistent homology as input for machine learning models, taking advantage of the ability to quantify multiscale topological features. In this talk, we will particularly introduce the application of persistent homology to machine learning potentials that predict the potential energy of materials using machine learning models, along with specific examples in amorphous materials.