Machine Learning of Coupled Cluster (T)-Energy Corrections via Delta (Δ)-Learning

Ruth, Marcel; Gerbig, Dennis; Schreiner, Peter R.

doi:10.1021/acs.jctc.2c00501

Cited by 27 publications

(25 citation statements)

References 98 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Alongside IBS-like reaction schemes, hybrid QM/ML delta machine learning (ΔML) techniques have emerged to correct for deficiencies in low-level electronic structure calculations. − DFT is capable of capturing a high fraction of the true molecular energy. However, there is a portion of the energy which can only be captured using the most sophisticated correlated methods. ,, Unfortunately, determining this portion of the energy can be prohibitive for larger molecules.…”

Section: Introductionmentioning

confidence: 99%

Quantitative Prediction of Vertical Ionization Potentials from DFT via a Graph-Network-Based Delta Machine Learning Model Incorporating Electronic Descriptors

2023

View full text Add to dashboard Cite

While accurate wave function theories like CCSD(T) are capable of modeling molecular chemical processes, the associated steep computational scaling renders them intractable for treating large systems or extensive databases. In contrast, density functional theory (DFT) is much more computationally feasible yet often fails to quantitatively describe electronic changes in chemical processes. Herein, we report an efficient delta machine learning (ΔML) model that builds on the Connectivity-Based Hierarchy (CBH) schemean error correction approach based on systematic molecular fragmentation protocolsand achieves coupled cluster accuracy on vertical ionization potentials by correcting for deficiencies in DFT. The present study integrates concepts from molecular fragmentation, systematic error cancellation, and machine learning. First, we show that by using an electron population difference map, ionization sites within a molecule may be readily identified, and CBH correction schemes for ionization processes may be automated. As a central feature of our work, we employ a graph-based QM/ML model, which embeds atom-centered features describing CBH fragments into a computational graph to further increase accuracy for the prediction of vertical ionization potentials. In addition, we show that the incorporation of electronic descriptors from DFT, namely electron population difference features, improves model performance well beyond chemical accuracy (1 kcal/mol) to approach benchmark accuracy. While the raw DFT results are strongly dependent on the underlying functional used, for our best models, the performance is robust and much less dependent on the functional.

show abstract

Section: Introductionmentioning

confidence: 99%

Quantitative Prediction of Vertical Ionization Potentials from DFT via a Graph-Network-Based Delta Machine Learning Model Incorporating Electronic Descriptors

2023

View full text Add to dashboard Cite

show abstract

“…186 As is common for the ML field, different flavours of Δ-ML exist. 146,170,172,173,181,182,186–189…”

Section: Knowledge Transfermentioning

confidence: 99%

Neural network potentials for chemistry: concepts, applications and prospects

et al. 2023

View full text Add to dashboard Cite

show abstract

“…In this context, the NN model is typically trained first with low-fidelity data from a low-level theory, and then retrained with high-fidelity data from a higher level of theory. Similarly, ∆-learning methods, relying on telescoping sums, and with connections to composite methods [52,53], have seen significant utilization in recent years [7,27,35,40,41,54]. In this context, a low-fidelity PES is typically fitted to low-fidelity data, and a discrepancy PES is fitted to the discrepancy (∆) between low and high fidelity data, such that the sum of the two PES predictions provides an accurate fit to high-fidelity data.…”

Section: Introductionmentioning

confidence: 99%

Multifidelity neural network formulations for prediction of reactive molecular potential energy surfaces

Yang

Eldred

Zádor

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper focuses on the development of multifidelity modeling approaches using neural network surrogates, where training data arising from multiple model forms and resolutions are integrated to predict high-fidelity response quantities of interest at lower cost. We focus on the context of quantum chemistry and the integration of information from multiple levels of theory. Important foundations include the use of symmetry function-based atomic energy vector constructions as feature vectors for representing structures across families of molecules, and single-fidelity neural network training capabilities that learn the relationships needed to map feature vectors to potential energy predictions. These foundations are embedded within several multifidelity topologies that decompose the high-fidelity mapping into model-based components, including sequential formulations that admit a general nonlinear mapping across fidelities and discrepancy-based formulations that presume an additive decomposition. Methodologies are first explored and demonstrated on a pair of simple analytical test problems, and then deployed for potential energy prediction for C5H5 using B2PLYPD3 6-311++G(d,p) for high fidelity simulation data and Hartree Fock 6-31G for low fidelity data. For the common case of limited access to high-fidelity data, our computational results demonstrate that multifidelity neural network potential energy surface constructions achieve roughly an order of magnitude improvement, either in terms of test error reduction for equivalent total simulation cost or reduction in total cost for equivalent error.

show abstract

Machine Learning of Coupled Cluster (T)-Energy Corrections via Delta (Δ)-Learning

Cited by 27 publications

References 98 publications

Quantitative Prediction of Vertical Ionization Potentials from DFT via a Graph-Network-Based Delta Machine Learning Model Incorporating Electronic Descriptors

Quantitative Prediction of Vertical Ionization Potentials from DFT via a Graph-Network-Based Delta Machine Learning Model Incorporating Electronic Descriptors

Neural network potentials for chemistry: concepts, applications and prospects

Multifidelity neural network formulations for prediction of reactive molecular potential energy surfaces

Contact Info

Product

Resources

About