Electronic spectra from TDDFT and machine learning in chemical space.

  title={Electronic spectra from TDDFT and machine learning in chemical space.},
  author={Raghunathan Ramakrishnan and Mia Hartmann and Enrico Tapavicza and O. Anatole von Lilienfeld},
  journal={The Journal of chemical physics},
  volume={143 8},
Due to its favorable computational efficiency, time-dependent (TD) density functional theory (DFT) enables the prediction of electronic spectra in a high-throughput manner across chemical space. Its predictions, however, can be quite inaccurate. We resolve this issue with machine learning models trained on deviations of reference second-order approximate coupled-cluster (CC2) singles and doubles spectra from TDDFT counterparts, or even from DFT gap. We applied this approach to low-lying singlet… 

Figures from this paper

Electronic Descriptors for Supervised Spectroscopic Predictions
Electronic descriptors obtained from low-cost DFT methods are used to model the absorption spectra of organic molecules and it is demonstrated that with this electronic descriptors and the use of Neural Networks the authors can predict not only a density of excited states, but also getting very good estimation of the absorption spectrum and charge-transfer character of the electronic excited states.
On application of deep learning to simplified quantum-classical dynamics in electronically excited states
Deep learning (DL) is applied to simulate non-adiabatic molecular dynamics of phenanthrene, using the time-dependent density functional based tight binding (TD-DFTB) approach for excited states
Chemical diversity in molecular orbital energy predictions with kernel ridge regression.
This work investigates 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.
Capturing intensive and extensive DFT/TDDFT molecular properties with machine learning
Current state-of-the-art machine learning techniques are more suited to predict extensive as opposed to intensive quantities, but it is speculated on the need to develop global descriptors that can describe both extensive and intensive properties on equal footing.
Benchmarking DFTB+ on the first singlet vertical excitation energy of small organic molecules
DFTB+ is a popular and open source scientific software package that offers a fast and scalable implementation of the density functional based tight binding (DFTB) method. Among its multiple implemented
Multi-fidelity prediction of molecular optical peaks with deep learning
Optical properties are central to molecular design for many applications, including solar cells and biomedical imaging. A variety of ab initio and statistical methods have been developed for their
UV-Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning.
Predicting UV-visible absorption spectra is essential to understand photochemical processes and design energy materials. Quantum chemical methods can deliver accurate calculations of UV-visible
Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited.
The CQML model is a generalized unified recursive kernel ridge regression that exploits correlations implicitly encoded in training data composed of multiple levels in multiple dimensions, and suggests a possibly favorable trade-off between various hierarchical approximations whose computational cost scales differently with electron number.
Selected Machine Learning of HOMO-LUMO gaps with Improved Data-Efficiency
The findings suggest that selected QML, e.g. based on simple classifications prior to training, could help to successfully tackle challenging quantum property screening tasks of large libraries with high fidelity and low computational burden.
Machine learning model for non-equilibrium structures and energies of simple molecules.
The optimized structures and energies of the molecules are found to be comparable with DFT optimized structures, energies, and forces and offers the possibility to use a trained ML model to perform a classical simulation such as MMC without using any force field, thereby improving the accuracy of the simulation at low computational cost.


Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach.
The transferability of the approach is demonstrated, using semiempirical quantum chemistry and machine learning models trained on 1 and 10% of 134k organic molecules, to reproduce enthalpies of all remaining molecules at density functional theory level of accuracy.
Extensive TD-DFT Benchmark: Singlet-Excited States of Organic Molecules.
It turned out that CC2 and TD-DFT errors are of the same order of magnitude, once the above-mentioned hybrids are selected, and the M05 and CAM-B3LYP schemes deliver fairly small deviations but do not outperform standard hybrids such as X3LYp or PBE0, at least within the vertical approximation.
Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space
A systematic hierarchy of efficient empirical methods to estimate atomization and total energies of molecules and is achieved by a vectorized representation of molecules (so-called Bag of Bonds model) that exhibits strong nonlocality in chemical space.
Assessing Excited State Methods by Adiabatic Excitation Energies.
We compile a 109-membered benchmark set of adiabatic excitation energies (AEEs) from high-resolution gas-phase experiments. Our data set includes a variety of organic chromophores with up to 46
How to represent crystal structures for machine learning: Towards fast prediction of electronic properties
It is found that conventional representations of the input data, such as the Coulomb matrix, are not suitable for the training of learning machines in the case of periodic solids and proposes a novel crystal structure representation for which learning and competitive prediction accuracies become possible within an unrestricted class of spd systems of arbitrary unit-cell size.
Many Molecular Properties from One Kernel in Chemical Space.
Property-independent kernels for machine learning models of arbitrarily many molecular properties enable the instantaneous generation of machine learning Models which can be systematically improved through the addition of more data.
CC2 excitation energy calculations on large molecules using the resolution of the identity approximation
A new implementation of the approximate coupled cluster singles and doubles method CC2 is reported, which is suitable for large scale integral-direct calculations. It employs the resolution of the
Excitation energies in density functional theory: an evaluation and a diagnostic test.
The study highlights the ambiguous nature of the term "charge transfer," providing insight into the observation that while many charge-transfer excitations are poorly described by GGA and hybrid functionals, others are accurately reproduced.
Fast and accurate modeling of molecular atomization energies with machine learning.
A machine learning model is introduced to predict atomization energies of a diverse set of organic molecules, based on nuclear charges and atomic positions only, and applicability is demonstrated for the prediction of molecular atomization potential energy curves.
Assessment of the Perdew–Burke–Ernzerhof exchange-correlation functional
In order to discriminate between approximations to the exchange-correlation energy EXC[ρ↑,ρ↓], we employ the criterion of whether the functional is fitted to a certain experimental data set or if it