Prediction of Stress-Strain Curve of Block Copolymers Using Transfer Learning of 3D Convolutional Neural Network.

• DOI: 10.2477/jccj.2021-0037
• Publication Date: 2021-11-29T22:10:01Z
• Authors: Hironobu KAWAGUCHI; Mariko I. ITO; Takeshi AOYAGI; Takaaki OHNISHI
• Keywords: Transfer of learning; Stress–strain curve; Strain (injury)

The simulation to obtain stress-strain curves of block copolymers requires large computational resources. As an alternative simulation, a method for high-throughput prediction using three-dimensional convolutional neural network has been reported. In this study, we incorporated shortcut coupling into the and performed pre-training transfer learning in step-by-step manner successfully predict curve with high accuracy while reducing training cost.