Publication | Open Access

Scalable training of graph convolutional neural networks for fast and accurate predictions of HOMO-LUMO gap in molecules