Fast and Accurate Machine Learning Prediction of Phonon Scattering Rates and Lattice Thermal Conductivity

Published in npj Computational Materials, 2023

Built the first machine learning model that can predict phonon scattering rates and thermal conductivity at the experimental and first principles accuracy level, with up to two orders of magnitude acceleration.
Trained deep neural network using TensorFlow. Mitigated challenges associated with the high skewness of phonon scattering rates and their complex contributions to the total thermal resistance. Performed transfer learning to further improve model performance.