Abstract:Recently, a novel, neural-networks-based method, the DFT-NEURON method, was developed to improve the accuracy of first-principles calculations and was applied to correct the systematic deviations of the calculated heats of formation for small-to-medium-sized organic molecules (Hu, L. H.; Wang, X. J.; Wong, L. H.; Chen, G. H. J. Chem. Phys. 2003, 119, 11501). In this work, we examine its theoretical foundation and generalize it to adopt any other statistical correction approaches, in particular, the multiple linear regression method. Both neural-networks-based and multiple-linear-regression-based correction approaches are applied to calculate the Gibbs energies of formation, ionization energies, electron affinities, and absorption energies of small-to-medium-sized molecules and lead to greatly improved calculation results as compared to the conventional first-principles methods. For instance, after the neural networks correction (multiple linear regression correction), the root-mean-square (RMS) deviations of the calculated standard Gibbs energy of formation for 180 organic molecules are reduced from 12.5, 13.8, and 22.3 kcal(.)mol(-1) to 4.7 (5.4), 3.2 (3.5), and 3.0 (3.2) kcal(.)mol(-1) for B3LYP/6-31G(d), B3LYP/6-311+G(3df,2p), and B3LYP/6-311+G(d,p) calculations, respectively, and the RMS deviation of the calculated absorption energies of 60 organic molecules is reduced from 0.33 eV to 0.09 (0.14) eV for the TDDFT/B3LYP/6-31G(d) calculation. In general, the neural networks correction approach leads to better results than the multiple linear regression correction approach. All these demonstrate that the statistical-correction-based first-principles calculations yield excellent results and may be employed routinely as predictive tools in materials research and design.

An accurate density functional theory calculation for electronic excitation energies: the least-squares support vector machine.

Improving the Accuracy of Low Level Density Functional Theory Calculation for Absorption Energies: the Least Squares Support Vector Machine

Improving the Accuracy of Low Level Quantum Chemical Calculation for Absorption Energies Based on Least Squares Support Vector Machine

Density Functional Prediction of Quasiparticle, Excitation, and Resonance Energies of Molecules with a Global Scaling Correction Approach

Accurate Prediction of Transition Energies in Organic Molecules

A semilocal machine-learning correction to density functional approximations

Improving The Accuracy Of Density-Functional Theory Calculation: The Genetic Algorithm And Neural Network Approach

Density Functionals with Quantum Chemical Accuracy: From Machine Learning to Molecular Dynamics

A New Fragment-Based Approach For Calculating Electronic Excitation Energies Of Large Systems

Machine Learning Methods to Predict Density Functional Theory B3LYP Energies of HOMO and LUMO Orbitals

Density functional approximations for charge transfer excitations with intermediate spatial overlap.

Improving the accuracy of density-functional theory calculation: The statistical correction approach

Machine learning for accuracy in density functional approximations

The Calculation of Excitation Energies Based on the Relativistic Two-Component Zeroth-Order Regular Approximation and Time-Dependent Density-Functional with Full Use of Symmetry

Approximating Quasiparticle and Excitation Energies from Ground State Generalized Kohn-Sham Calculations.

Density Functional Calculations On First-Row Transition Metals

Machine-learning prediction of exciton binding energies in two-dimensional materials for photonics application

Improving Density Functional Prediction of Molecular Thermochemical Properties with a Machine-Learning-Corrected Generalized Gradient Approximation.

Machine Learning Dielectric Screening for the Simulation of Excited State Properties of Molecules and Materials

Accuracy of Charge Densities in Electronic Structure Calculations

Achieving vibrational energies of diatomic systems with high quality by machine learning improved DFT method