Symmetry-invariant quantum machine learning force fields
Machine learning techniques are essential tools to compute efficient, yet accurate, force fields for atomistic simulations. This approach has recently been extended to incorporate quantum computational methods, making use of variational quantum learning models to predict potential energy surfaces an...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2025-01-01
|
| Series: | New Journal of Physics |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/1367-2630/adad0c |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823858336845529088 |
|---|---|
| author | Isabel Nha Minh Le Oriel Kiss Julian Schuhmacher Ivano Tavernelli Francesco Tacchino |
| author_facet | Isabel Nha Minh Le Oriel Kiss Julian Schuhmacher Ivano Tavernelli Francesco Tacchino |
| author_sort | Isabel Nha Minh Le |
| collection | DOAJ |
| description | Machine learning techniques are essential tools to compute efficient, yet accurate, force fields for atomistic simulations. This approach has recently been extended to incorporate quantum computational methods, making use of variational quantum learning models to predict potential energy surfaces and atomic forces from ab initio training data. However, the trainability and scalability of such models are still limited, due to both theoretical and practical barriers. Inspired by recent developments in geometric classical and quantum machine learning, here we design quantum neural networks that explicitly incorporate, as a data-inspired prior, an extensive set of physically relevant symmetries. We find that our invariant quantum learning models outperform their more generic counterparts on individual molecules of growing complexity. Furthermore, we study a water dimer as a minimal example of a system with multiple components, showcasing the versatility of our proposed approach and opening the way towards larger simulations. Finally, we perform a barren plateau analysis and numerically observe that our model does not exhibit a barren plateau in the shallow depth regime. Our results suggest that molecular force fields generation can significantly profit from leveraging the framework of geometric quantum machine learning, and that chemical systems represent, in fact, an interesting and rich playground for the development and application of advanced quantum machine learning tools. |
| format | Article |
| id | doaj-art-56fd336986464c53a6c69cea483a1a8e |
| institution | Kabale University |
| issn | 1367-2630 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | New Journal of Physics |
| spelling | doaj-art-56fd336986464c53a6c69cea483a1a8e2025-02-11T11:38:36ZengIOP PublishingNew Journal of Physics1367-26302025-01-0127202301510.1088/1367-2630/adad0cSymmetry-invariant quantum machine learning force fieldsIsabel Nha Minh Le0https://orcid.org/0000-0001-6707-044XOriel Kiss1https://orcid.org/0000-0001-7461-3342Julian Schuhmacher2https://orcid.org/0000-0002-7011-6477Ivano Tavernelli3https://orcid.org/0000-0001-5690-1981Francesco Tacchino4https://orcid.org/0000-0003-2008-5956IBM Quantum , IBM Research Europe—Zurich, 8803 Rueschlikon, Switzerland; Institute for Quantum Information , RWTH Aachen University, 52074 Aachen, Germany; Department of Computer Science, Technical University of Munich , School of Computation, Information and Technology, 85748 Garching, GermanyEuropean Organization for Nuclear Research (CERN) , 1211 Geneva, Switzerland; Department of Nuclear and Particle Physics, University of Geneva , 1211 Geneva, SwitzerlandIBM Quantum , IBM Research Europe—Zurich, 8803 Rueschlikon, SwitzerlandIBM Quantum , IBM Research Europe—Zurich, 8803 Rueschlikon, SwitzerlandIBM Quantum , IBM Research Europe—Zurich, 8803 Rueschlikon, SwitzerlandMachine learning techniques are essential tools to compute efficient, yet accurate, force fields for atomistic simulations. This approach has recently been extended to incorporate quantum computational methods, making use of variational quantum learning models to predict potential energy surfaces and atomic forces from ab initio training data. However, the trainability and scalability of such models are still limited, due to both theoretical and practical barriers. Inspired by recent developments in geometric classical and quantum machine learning, here we design quantum neural networks that explicitly incorporate, as a data-inspired prior, an extensive set of physically relevant symmetries. We find that our invariant quantum learning models outperform their more generic counterparts on individual molecules of growing complexity. Furthermore, we study a water dimer as a minimal example of a system with multiple components, showcasing the versatility of our proposed approach and opening the way towards larger simulations. Finally, we perform a barren plateau analysis and numerically observe that our model does not exhibit a barren plateau in the shallow depth regime. Our results suggest that molecular force fields generation can significantly profit from leveraging the framework of geometric quantum machine learning, and that chemical systems represent, in fact, an interesting and rich playground for the development and application of advanced quantum machine learning tools.https://doi.org/10.1088/1367-2630/adad0cmolecular force fieldsgeometric quantum machine learningequivariant quantum neural networks |
| spellingShingle | Isabel Nha Minh Le Oriel Kiss Julian Schuhmacher Ivano Tavernelli Francesco Tacchino Symmetry-invariant quantum machine learning force fields New Journal of Physics molecular force fields geometric quantum machine learning equivariant quantum neural networks |
| title | Symmetry-invariant quantum machine learning force fields |
| title_full | Symmetry-invariant quantum machine learning force fields |
| title_fullStr | Symmetry-invariant quantum machine learning force fields |
| title_full_unstemmed | Symmetry-invariant quantum machine learning force fields |
| title_short | Symmetry-invariant quantum machine learning force fields |
| title_sort | symmetry invariant quantum machine learning force fields |
| topic | molecular force fields geometric quantum machine learning equivariant quantum neural networks |
| url | https://doi.org/10.1088/1367-2630/adad0c |
| work_keys_str_mv | AT isabelnhaminhle symmetryinvariantquantummachinelearningforcefields AT orielkiss symmetryinvariantquantummachinelearningforcefields AT julianschuhmacher symmetryinvariantquantummachinelearningforcefields AT ivanotavernelli symmetryinvariantquantummachinelearningforcefields AT francescotacchino symmetryinvariantquantummachinelearningforcefields |