| Abstract Scope |
Machine learning is trending to be an integral part of thermal science. In this talk, I will introduce our efforts in utilizing machine learning (ML) techniques to predict phonon properties and thermal modeling. I will talk about the use of a new ML method, called transfer learning, to establish accurate models based on limited data for predicting phonon and thermal transport properties, such as phonon frequency gap, heat capacity, speed of sound and lattice thermal conductivity. |