README.md

ML-based semi-empirical model based-on the extended Hückel method (ML-EHM)

This repository contains the EHM-ML data

If you use the ML-EHM data or model please cite this paper:

Machine Learned Hückel Theory: Interfacing Physics and Deep Neural Networks:

Tetiana Zubatiuk, Benjamin Nebgen, Nicholas Lubbers, Justin S. Smith, Roman Zubatyuk, Guoqing Zhou, Christopher Koh, Kipton Barros, Olexandr Isayev, Sergei Tretiak. Machine Learned Hückel Theory: Interfacing Physics and Deep Neural Networks. arXiv:1909.12963, (https://arxiv.org/abs/1909.12963)

More detailed information about COMP6 benchmark(https://github.com/isayev/COMP6), its design and composion can be found in the following publicaiton:

Justin S. Smith, Ben Nebgen, Nicholas Lubbers, Olexandr Isayev, Adrian E. Roitberg. Less is more: sampling chemical space with active learning. The Journal of Chemical Physics 148, 241733 (2018), (https://aip.scitation.org/doi/abs/10.1063/1.5023802)