[S1] S1 New Materials Science on Nanoscale Structures and Functions at Crystal Interface II

1:40 PM - 2:00 PM JST(4:40 AM - 5:00 AM UTC)

[S1.18] Discovery of High-Dielectric Materials Using a Factorized Machine Learning Model

○Masamune Kawano^1,2, Atsushi Takigawa^1,2, Shin Kiyohara², Yu Kumagai² (1. School of Engineering, Tohoku University, 2. Institute for Materials Research, Tohoku University)

Keywords:

High-dielectric materials,Machine learning

This study applies a factorized machine learning model to 3,034 nonmetallic compounds and validates its predictions of ionic dielectric constants. By analyzing large-error cases, we identify phonon-prediction errors as the main source and present methods to improve accuracy and their effects.