Local Binary and Multiclass SVMs Trained on a Quantum Annealer
Support vector machines (SVMs) are widely used machine learning models, with formulations for both classification and regression tasks. In the last years, with the advent of working quantum annealers, hybrid SVM models characterized by quantum training and classical execution have been introduced. T...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Transactions on Quantum Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10706813/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832586899268567040 |
|---|---|
| author | Enrico Zardini Amer Delilbasic Enrico Blanzieri Gabriele Cavallaro Davide Pastorello |
| author_facet | Enrico Zardini Amer Delilbasic Enrico Blanzieri Gabriele Cavallaro Davide Pastorello |
| author_sort | Enrico Zardini |
| collection | DOAJ |
| description | Support vector machines (SVMs) are widely used machine learning models, with formulations for both classification and regression tasks. In the last years, with the advent of working quantum annealers, hybrid SVM models characterized by quantum training and classical execution have been introduced. These models have demonstrated comparable performance to their classical counterparts. However, they are limited in the training set size due to the restricted connectivity of the current quantum annealers. Hence, to take advantage of large datasets, a strategy is required. In the classical domain, local SVMs, namely, SVMs trained on the data samples selected by a <inline-formula><tex-math notation="LaTeX">$k$</tex-math></inline-formula>-nearest neighbors model, have already proven successful. Here, the local application of quantum-trained SVM models is proposed and empirically assessed. In particular, this approach allows overcoming the constraints on the training set size of the quantum-trained models while enhancing their performance. In practice, the fast local kernel support vector machine (FaLK-SVM) method, designed for efficient local SVMs, has been combined with quantum-trained SVM models for binary and multiclass classification. In addition, for comparison, FaLK-SVM has been interfaced for the first time with a classical single-step multiclass SVM model. Concerning the empirical evaluation, D-Wave's quantum annealers and real-world datasets taken from the remote sensing domain have been employed. The results have shown the effectiveness and scalability of the proposed approach, but also its practical applicability in a real-world large-scale scenario. |
| format | Article |
| id | doaj-art-f3d02ddba02d4efb952aeb4dd37c95c9 |
| institution | Kabale University |
| issn | 2689-1808 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Quantum Engineering |
| spelling | doaj-art-f3d02ddba02d4efb952aeb4dd37c95c92025-01-25T00:03:46ZengIEEEIEEE Transactions on Quantum Engineering2689-18082024-01-01511210.1109/TQE.2024.347587510706813Local Binary and Multiclass SVMs Trained on a Quantum AnnealerEnrico Zardini0https://orcid.org/0000-0002-7475-7183Amer Delilbasic1https://orcid.org/0000-0001-7845-5193Enrico Blanzieri2https://orcid.org/0000-0001-6524-0601Gabriele Cavallaro3https://orcid.org/0000-0002-3239-9904Davide Pastorello4https://orcid.org/0000-0001-5915-6796Department of Information Engineering and Computer Science, University of Trento, Trento, ItalyJülich Supercomputing Centre, Wilhelm-Johnen Straße, Jülich, GermanyDepartment of Information Engineering and Computer Science, University of Trento, Trento, ItalyJülich Supercomputing Centre, Wilhelm-Johnen Straße, Jülich, GermanyTrento Institute for Fundamental Physics and Applications, Trento, ItalySupport vector machines (SVMs) are widely used machine learning models, with formulations for both classification and regression tasks. In the last years, with the advent of working quantum annealers, hybrid SVM models characterized by quantum training and classical execution have been introduced. These models have demonstrated comparable performance to their classical counterparts. However, they are limited in the training set size due to the restricted connectivity of the current quantum annealers. Hence, to take advantage of large datasets, a strategy is required. In the classical domain, local SVMs, namely, SVMs trained on the data samples selected by a <inline-formula><tex-math notation="LaTeX">$k$</tex-math></inline-formula>-nearest neighbors model, have already proven successful. Here, the local application of quantum-trained SVM models is proposed and empirically assessed. In particular, this approach allows overcoming the constraints on the training set size of the quantum-trained models while enhancing their performance. In practice, the fast local kernel support vector machine (FaLK-SVM) method, designed for efficient local SVMs, has been combined with quantum-trained SVM models for binary and multiclass classification. In addition, for comparison, FaLK-SVM has been interfaced for the first time with a classical single-step multiclass SVM model. Concerning the empirical evaluation, D-Wave's quantum annealers and real-world datasets taken from the remote sensing domain have been employed. The results have shown the effectiveness and scalability of the proposed approach, but also its practical applicability in a real-world large-scale scenario.https://ieeexplore.ieee.org/document/10706813/Localityquantum annealing (QA)quantum computingsupport vector machines (SVMs) |
| spellingShingle | Enrico Zardini Amer Delilbasic Enrico Blanzieri Gabriele Cavallaro Davide Pastorello Local Binary and Multiclass SVMs Trained on a Quantum Annealer IEEE Transactions on Quantum Engineering Locality quantum annealing (QA) quantum computing support vector machines (SVMs) |
| title | Local Binary and Multiclass SVMs Trained on a Quantum Annealer |
| title_full | Local Binary and Multiclass SVMs Trained on a Quantum Annealer |
| title_fullStr | Local Binary and Multiclass SVMs Trained on a Quantum Annealer |
| title_full_unstemmed | Local Binary and Multiclass SVMs Trained on a Quantum Annealer |
| title_short | Local Binary and Multiclass SVMs Trained on a Quantum Annealer |
| title_sort | local binary and multiclass svms trained on a quantum annealer |
| topic | Locality quantum annealing (QA) quantum computing support vector machines (SVMs) |
| url | https://ieeexplore.ieee.org/document/10706813/ |
| work_keys_str_mv | AT enricozardini localbinaryandmulticlasssvmstrainedonaquantumannealer AT amerdelilbasic localbinaryandmulticlasssvmstrainedonaquantumannealer AT enricoblanzieri localbinaryandmulticlasssvmstrainedonaquantumannealer AT gabrielecavallaro localbinaryandmulticlasssvmstrainedonaquantumannealer AT davidepastorello localbinaryandmulticlasssvmstrainedonaquantumannealer |