Multitask methods for predicting molecular properties from heterogeneous data
Name
014114_1_5.0201681.pdf
Description
Published version
Size
7.14 MB
Format
Adobe PDF
Checksum (MD5)
690223f0311ba191b611668976a41ce7
Author(s)
Fisher, KE
Herbst, MF
Marzouk, YM
Date Issued
July 3, 2024
Journal
The Journal of Chemical Physics
Publisher
AIP Publishing
Citation
K. E. Fisher, M. F. Herbst, Y. M. Marzouk; Multitask methods for predicting molecular properties from heterogeneous data. J. Chem. Phys. 7 July 2024; 161 (1): 014114.
Version
Final published version
Abstract
Data generation remains a bottleneck in training surrogate models to predict molecular properties. We demonstrate that multitask Gaussian process regression overcomes this limitation by leveraging both expensive and cheap data sources. In particular, we consider training sets constructed from coupled-cluster (CC) and density functional theory (DFT) data. We report that multitask surrogates can predict at CC-level accuracy with a reduction in data generation cost by over an order of magnitude. Of note, our approach allows the training set to include DFT data generated by a heterogeneous mix of exchange–correlation functionals without imposing any artificial hierarchy on functional accuracy. More generally, the multitask framework can accommodate a wider range of training set structures—including the full disparity between the different levels of fidelity—than existing kernel approaches based on Δ-learning although we show that the accuracy of the two approaches can be similar. Consequently, multitask regression can be a tool for reducing data generation costs even further by opportunistically exploiting existing data sources.
MIT Department
Massachusetts Institute of Technology. Department of Aeronautics and Astronautics
Terms of Use
Creative Commons Attribution
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1063/5.0201681
