[17p-P08-1~8] 23.1 Joint Session N "Informatics"

[17p-P08-6] Generation of high molecular three-dimentional structure with variational auto encorder

〇(M1)Akihiro Matsunaga¹, Ichiki Akihisa² (1.Nagoya Univ., 2.Innovation for Future Society Inst.)

Keywords:machine learning, high molecular

Differences in microstructure of polymer materials have a great effect on their physical properties. As a result, it is difficult to predict the physical properties from the molecular structure, so it is necessary to create a microstructure by molecular dynamics calculation, but a large-scale simulation takes an enormous amount of time. In this study, we created a model that generates polymer microstructures with different molecular structures from polymer structures by coarse-vision molecular dynamics simulation using a machine learning method called Variational Auto-Encoder.