We present a machine learning (ML) framework that predicts G0W0 quasiparticle energies across molecular dynamics (MD) trajectories with high accuracy and efficiency. Using only DFT-derived mean-field eigenvalues and exchange-correlation potentials, the model is trained on 25\% of MD snapshots and achieves RMSEs below 0.1 eV. It accurately reproduces k-resolved quasiparticle band structures and density of states, even for BN polymorphs excluded from the training data. This approach bypasses the computational bottlenecks of G0W0 simulations over dynamic configurations, offering a scalable route to excited-state electronic structure simulations with many-body accuracy.
@article{arxiv.2505.02421,
title = {High-Throughput GW Calculations via Machine Learning},
author = {Ragab. A. Abdelghany and Chih-En Hsu and Hung-Chung Hsueh and Yuan-Hong Tsai and Ming-Chiang Chung},
journal= {arXiv preprint arXiv:2505.02421},
year = {2025}
}