A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
Chemical diversity in molecular orbital energy predictions with kernel ridge regression
Tekijät: Stuke A, Todorovic M, Rupp M, Kunkel C, Ghosh K, Himanen L, Rinke P
Kustantaja: AMER INST PHYSICS
Julkaisuvuosi: 2019
Journal: Journal of Chemical Physics
Tietokannassa oleva lehden nimi: JOURNAL OF CHEMICAL PHYSICS
Lehden akronyymi: J CHEM PHYS
Artikkelin numero: 204121
Vuosikerta: 150
Numero: 20
Sivujen määrä: 13
ISSN: 0021-9606
DOI: https://doi.org/10.1063/1.5086105
Tiivistelmä
Instant machine learning predictions of molecular properties are desirable for materials design, but the predictive power of the methodology is mainly tested on well-known benchmark datasets. Here, we investigate the performance of machine learning with kernel ridge regression (KRR) for the prediction of molecular orbital energies on three large datasets: the standard QM9 small organic molecules set, amino acid and dipeptide conformers, and organic crystal-forming molecules extracted from the Cambridge Structural Database. We focus on the prediction of highest occupied molecular orbital (HOMO) energies, computed at the density-functional level of theory. Two different representations that encode the molecular structure are compared: the Coulomb matrix (CM) and the many-body tensor representation (MBTR). We find that KRR performance depends significantly on the chemistry of the underlying dataset and that the MBTR is superior to the CM, predicting HOMO energies with a mean absolute error as low as 0.09 eV. To demonstrate the power of our machine learning method, we apply our model to structures of 10k previously unseen molecules. We gain instant energy predictions that allow us to identify interesting molecules for future applications. Published under license by AlP Publishing.
Instant machine learning predictions of molecular properties are desirable for materials design, but the predictive power of the methodology is mainly tested on well-known benchmark datasets. Here, we investigate the performance of machine learning with kernel ridge regression (KRR) for the prediction of molecular orbital energies on three large datasets: the standard QM9 small organic molecules set, amino acid and dipeptide conformers, and organic crystal-forming molecules extracted from the Cambridge Structural Database. We focus on the prediction of highest occupied molecular orbital (HOMO) energies, computed at the density-functional level of theory. Two different representations that encode the molecular structure are compared: the Coulomb matrix (CM) and the many-body tensor representation (MBTR). We find that KRR performance depends significantly on the chemistry of the underlying dataset and that the MBTR is superior to the CM, predicting HOMO energies with a mean absolute error as low as 0.09 eV. To demonstrate the power of our machine learning method, we apply our model to structures of 10k previously unseen molecules. We gain instant energy predictions that allow us to identify interesting molecules for future applications. Published under license by AlP Publishing.
Turun yliopiston Tutkimustietojärjestelmän portaali