Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

International Conference on Information, Implementation, and Innovation in Technology (I2ITCON)

Optimizing the Quantum Circuit of Quantum SVM Using Hybrid Quantum Natural Gradient Descent(QNGD) and Quantum Butterfly Optimization Algorithm(QBOA)

Raakesh Premkumar, T.B Sivakumar, L. Maria Michael Visuwasam, G. Yuyaraj, Tharun Kumar C

Published Jul 4, 2025• 0 citations
Publisher

Abstract

This research endeavor is centered on the enhancement of the quantum circuit associated with the Quantum Support Vector Machine (QSVM), a fundamental algorithm within the realm of quantum machine learning, through the application of a hybrid optimization methodology that amalgamates Quantum Natural Gradient Descent (QNGD) with the Quantum Butterfly Optimization Algorithm (QBOA). The QSVM exploits quantum computational capabilities for proficient classification tasks within elevated-dimensional frameworks; however, its operational efficacy is frequently compromised by non-ideal quantum circuit parameters and the deleterious effects of noise present in near-term quantum computational devices. The suggested hybrid optimization technique augments the performance of QSVM by fusing the localized refinement capabilities of QNGD, which employs the Fisher Information Matrix for effective parameter adjustments, with the expansive exploration potential of QBOA, inspired by the foraging behaviors of butterflies. This collaborative approach yields accelerated convergence (38 iterations), enhanced classification accuracy ($\mathbf{9 6. 8 \%}$), diminished circuit complexity (44 quantum gates), and increased resilience to noise (fidelity:$\mathbf{0. 9 8 5}$). In comparison to traditional optimization techniques such as Differential Evolution (DE), Grey Wolf Optimizer (GWO), and Firefly Algorithm (FA), the proposed framework exhibits markedly superior efficacy in the refinement of quantum kernel computations, entanglement strategies, and the overall depth of the circuit while concurrently alleviating the impacts of noise.

DOI
10.1109/i2itcon65200.2025.11210661

Related papers

Optimizing Quantum Support Vector Machine (QSVM) Circuits Using Hybrid Quantum Natural Gradient Descent (QNGD) and Whale Optimization Algorithm (WOA)

Tamilselvan Arjunan, Gopalakrishnan Arjunan, N. Jagadish Kumar +1 more

This research endeavor centers on augmenting the quantum circuitry of the Quantum Support Vector Machine (QSVM), a pivotal algorithm in the realm of quantum machine learning, via a hybrid optimization methodology that amalgamates Quantum Natural Gradient Descent (QNGD) with the Whale Optimization Algorithm (WOA). QSVM harnesses the power of quantum computing to facilitate proficient classification within high-dimensional spaces; however, its efficacy is frequently compromised by suboptimal circuit parameters and quantum noise prevalent in near-term quantum devices. The proposed hybrid optimization paradigm integrates the global search proficiency of WOA with the effective parameter refinement capabilities of QNGD, culminating in expedited convergence (38 iterations), enhanced classification accuracy (96.8%), optimized circuit complexity (44 quantum gates), and increased resilience against noise (fidelity: 0.985). In comparison to conventional optimization techniques such as Differential Evolution (DE), Grey Wolf Optimizer (GWO), and Firefly Algorithm (FA), our methodology demonstrates superior efficacy by refining quantum kernel computation, entanglement configurations, and circuit depth, while concurrently alleviating the impacts of quantum noise.

International Conference for Emerging Technology (INCET)•May 23, 2025•2 citations

Optimizing the Quantum Circuit of Quantum K-Nearest Neighbors (QKNN) Using Hybrid Gradient Descent and Golden Eagle Optimization Algorithm

Tamilselvan Arjunan, Gopalakrishnan Arjunan, N. Jagadish Kumar +1 more

This study concentrates on the enhancement of the quantum circuit pertaining to Quantum K-Nearest Neighbors (QKNN), which is a fundamental algorithm within the domain of quantum machine learning, through the implementation of a hybrid optimization methodology that integrates Gradient Descent with the Golden Eagle Optimization Algorithm (GEOA). QKNN harnesses the capabilities of quantum computing to process high-dimensional datasets with enhanced computational efficiency and precision; however, its effectiveness is frequently compromised by inadequate quantum circuit parameters and elevated noise levels present in near-term quantum devices. The proposed hybrid optimization methodology augments Quantum K-Nearest Neighbors (QKNN) through the amalgamation of the global search capabilities inherent in Golden Eagle Optimization (GEO) and the localized refinement efficacy provided by Gradient Descent (GD). This collaborative interplay culminates in expedited convergence (40 iterations), elevated classification accuracy (97.2 %), diminished circuit complexity (42 quantum gates), and enhanced resilience against gate noise (fidelity: 0.983). In contrast to conventional optimization techniques such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA), our proposed methodology demonstrates superior efficacy by optimizing quantum distance computation, entanglement strategies, and circuit depth, while simultaneously attenuating the effects of gate noise.

International Conference on Computing Technologies & Data Communication (ICCTDC)•Jul 4, 2025•3 citations

Optimizing Quantum Decision Trees (QDT) Circuits Using Hybrid Golden Eagle Optimization with Adaptive Gradient Descent

L. Maria Michael Visuwasam, Sudha B, G. Kalpana +1 more

This research centers on enhancing quantum circuits for Quantum Decision Trees (QDT) a highly promising algorithm in the realm of quantum machine learning aimed at classification and pattern recognition-via a hybrid metaheuristic optimization framework that integrates Golden Eagle Optimization (GEO) with Adaptive Gradient Descent (AGD). Although QDTs provide interpretability and quantum-enhanced classification capabilities for high-dimensional data sets, their effectiveness is frequently limited by subpar circuit configurations, excessive gate depth, and susceptibility to quantum noise. The suggested hybrid approach utilizes the global search capability of GEO and the local convergence accuracy of AGD to optimize QDT parameters such as entanglement structure, placement of quantum gates, and decision thresholds. Experimental findings indicate that the proposed technique achieves quicker convergence (38 iterations), superior classification accuracy (97.8 %), decreased quantum circuit complexity (39 quantum gates), and enhanced noise resilience (fidelity: 0.986), surpassing traditional methods including Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA). These results affirm the promise of hybrid metaheuristic optimizers in improving quantum circuit efficiency for real-world quantum machine learning applications.

International Conference on Emerging Trends and Innovations in ICT (ICEI)•Jan 9, 2026•0 citations

Flexible Genetic Algorithm for Quantum Support Vector Machines

Nguyen Minh Duc, Vu Tuan Hai, Le Bin Ho +1 more

Quantum Support Vector Machines (QSVM) is one of the most promising frameworks in quantum machine learning, yet their performance depends on the design of the feature map. Conventional approaches rely on fixed quantum circuits, which often fail to generalize across datasets. To address this limitation, we propose GA-QSVM, a hybrid framework that employs Genetic Algorithms (GA) to automatically optimize feature maps. The proposed method introduces a configurable framework that flexibly defines the evolutionary parameters, enabling the construction of adaptive circuits. Experimental evaluation of datasets, including Digits, Fashion, Wine, and Breast Cancer, demonstrates that GA-QSVMs achieve a comparable accuracy compared to classical SVMs and standard QSVMs. Furthermore, transfer learning results indicate that GA-QSVM's circuits generalize effectively across datasets. These findings highlight the potential of evolutionary strategies to automate and enhance kernel design for future quantum machine learning applications.

quant-pharXiv:2511.19160•Machine Learning Science and Technology•Nov 25, 2025•0 citations

Machine Learning in the Quantum Age: Quantum vs. Classical Support Vector Machines

Davut Emre Tasar, Kutan Koruyan, Ceren Öcal Taşar

This work endeavors to juxtapose the efficacy of machine learning algorithms within classical and quantum computational paradigms. Particularly, by emphasizing on Support Vector Machines (SVM), we scrutinize the classification prowess of classical SVM and Quantum Support Vector Machines (QSVM) operational on quantum hardware over the Iris dataset. The methodology embraced encapsulates an extensive array of experiments orchestrated through the Qiskit library, alongside hyperparameter optimization. The findings unveil that in particular scenarios, QSVMs extend a level of accuracy that can vie with classical SVMs, albeit the execution times are presently protracted. Moreover, we underscore that augmenting quantum computational capacity and the magnitude of parallelism can markedly ameliorate the performance of quantum machine learning algorithms. This inquiry furnishes invaluable insights regarding the extant scenario and future potentiality of machine learning applications in the quantum epoch. Colab: https://t.ly/QKuz0

arXiv:2310.10910•Oct 17, 2023•0 citations

Quantum Kernel-Based Classification: An Optimized Quantum Support Vector Classifier Approach

Xiaoman Liu

Quantum machine learning has emerged as a promising frontier, leveraging the principles of quantum computing to enhance classical algorithms. This paper proposes an optimized Quantum Support Vector Classifier (QSVC) for the classification of the MNIST dataset (a benchmark in handwritten digit recognition). We employ a Chebyshev Parameterized Quantum Circuit (PQC) with an increased number of qubits and layers, and utilize an existing projected quantum kernel method to effectively capture the complex patterns in the data. To enhance model robustness, data augmentation techniques are applied, and dimensionality reduction is performed using Principal Component Analysis (PCA). The proposed QSVC demonstrates competitive accuracy on both the training and testing datasets, highlighting the potential of quantum kernel methods in practical machine learning applications. A comparative analysis with classical Support Vector Machines (SVM) demonstrates the feasibility of quantum methods for high-dimensional classification tasks. Additionally, we investigate the impact of varying the number of qubits in the quantum circuit, providing in-depth insights into the scalability and performance trade-offs of quantum kernel methods.

IEEE International Conference on Frontiers Technology of Information and Computer (ICFTIC)•Dec 13, 2024•0 citations
This research endeavor is centered on the enhancement of the quantum circuit associated with the Quantum Support Vector Machine (QSVM), a fundamental algorithm within the realm of quantum machine learning, through the application of a hybrid optimization methodology that amalgamates Quantum Natural Gradient Descent (QNGD) with the Quantum Butterfly Optimization Algorithm (QBOA). The QSVM exploits quantum computational capabilities for proficient classification tasks within elevated-dimensional frameworks; however, its operational efficacy is frequently compromised by non-ideal quantum circuit parameters and the deleterious effects of noise present in near-term quantum computational devices. The suggested hybrid optimization technique augments the performance of QSVM by fusing the localized refinement capabilities of QNGD, which employs the Fisher Information Matrix for effective parameter adjustments, with the expansive exploration potential of QBOA, inspired by the foraging behaviors of butterflies. This collaborative approach yields accelerated convergence (38 iterations), enhanced classification accuracy (96.8%\mathbf{9 6. 8 \%}96.8%), diminished circuit complexity (44 quantum gates), and increased resilience to noise (fidelity:0.985\mathbf{0. 9 8 5}0.985). In comparison to traditional optimization techniques such as Differential Evolution (DE), Grey Wolf Optimizer (GWO), and Firefly Algorithm (FA), the proposed framework exhibits markedly superior efficacy in the refinement of quantum kernel computations, entanglement strategies, and the overall depth of the circuit while concurrently alleviating the impacts of noise.
This research endeavor centers on augmenting the quantum circuitry of the Quantum Support Vector Machine (QSVM), a pivotal algorithm in the realm of quantum machine learning, via a hybrid optimization methodology that amalgamates Quantum Natural Gradient Descent (QNGD) with the Whale Optimization Algorithm (WOA). QSVM harnesses the power of quantum computing to facilitate proficient classification within high-dimensional spaces; however, its efficacy is frequently compromised by suboptimal circuit parameters and quantum noise prevalent in near-term quantum devices. The proposed hybrid optimization paradigm integrates the global search proficiency of WOA with the effective parameter refinement capabilities of QNGD, culminating in expedited convergence (38 iterations), enhanced classification accuracy (96.8%), optimized circuit complexity (44 quantum gates), and increased resilience against noise (fidelity: 0.985). In comparison to conventional optimization techniques such as Differential Evolution (DE), Grey Wolf Optimizer (GWO), and Firefly Algorithm (FA), our methodology demonstrates superior efficacy by refining quantum kernel computation, entanglement configurations, and circuit depth, while concurrently alleviating the impacts of quantum noise.
This study concentrates on the enhancement of the quantum circuit pertaining to Quantum K-Nearest Neighbors (QKNN), which is a fundamental algorithm within the domain of quantum machine learning, through the implementation of a hybrid optimization methodology that integrates Gradient Descent with the Golden Eagle Optimization Algorithm (GEOA). QKNN harnesses the capabilities of quantum computing to process high-dimensional datasets with enhanced computational efficiency and precision; however, its effectiveness is frequently compromised by inadequate quantum circuit parameters and elevated noise levels present in near-term quantum devices. The proposed hybrid optimization methodology augments Quantum K-Nearest Neighbors (QKNN) through the amalgamation of the global search capabilities inherent in Golden Eagle Optimization (GEO) and the localized refinement efficacy provided by Gradient Descent (GD). This collaborative interplay culminates in expedited convergence (40 iterations), elevated classification accuracy (97.2 %), diminished circuit complexity (42 quantum gates), and enhanced resilience against gate noise (fidelity: 0.983). In contrast to conventional optimization techniques such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA), our proposed methodology demonstrates superior efficacy by optimizing quantum distance computation, entanglement strategies, and circuit depth, while simultaneously attenuating the effects of gate noise.
This research centers on enhancing quantum circuits for Quantum Decision Trees (QDT) a highly promising algorithm in the realm of quantum machine learning aimed at classification and pattern recognition-via a hybrid metaheuristic optimization framework that integrates Golden Eagle Optimization (GEO) with Adaptive Gradient Descent (AGD). Although QDTs provide interpretability and quantum-enhanced classification capabilities for high-dimensional data sets, their effectiveness is frequently limited by subpar circuit configurations, excessive gate depth, and susceptibility to quantum noise. The suggested hybrid approach utilizes the global search capability of GEO and the local convergence accuracy of AGD to optimize QDT parameters such as entanglement structure, placement of quantum gates, and decision thresholds. Experimental findings indicate that the proposed technique achieves quicker convergence (38 iterations), superior classification accuracy (97.8 %), decreased quantum circuit complexity (39 quantum gates), and enhanced noise resilience (fidelity: 0.986), surpassing traditional methods including Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Simulated Annealing (SA). These results affirm the promise of hybrid metaheuristic optimizers in improving quantum circuit efficiency for real-world quantum machine learning applications.
Quantum Support Vector Machines (QSVM) is one of the most promising frameworks in quantum machine learning, yet their performance depends on the design of the feature map. Conventional approaches rely on fixed quantum circuits, which often fail to generalize across datasets. To address this limitation, we propose GA-QSVM, a hybrid framework that employs Genetic Algorithms (GA) to automatically optimize feature maps. The proposed method introduces a configurable framework that flexibly defines the evolutionary parameters, enabling the construction of adaptive circuits. Experimental evaluation of datasets, including Digits, Fashion, Wine, and Breast Cancer, demonstrates that GA-QSVMs achieve a comparable accuracy compared to classical SVMs and standard QSVMs. Furthermore, transfer learning results indicate that GA-QSVM's circuits generalize effectively across datasets. These findings highlight the potential of evolutionary strategies to automate and enhance kernel design for future quantum machine learning applications.
This work endeavors to juxtapose the efficacy of machine learning algorithms within classical and quantum computational paradigms. Particularly, by emphasizing on Support Vector Machines (SVM), we scrutinize the classification prowess of classical SVM and Quantum Support Vector Machines (QSVM) operational on quantum hardware over the Iris dataset. The methodology embraced encapsulates an extensive array of experiments orchestrated through the Qiskit library, alongside hyperparameter optimization. The findings unveil that in particular scenarios, QSVMs extend a level of accuracy that can vie with classical SVMs, albeit the execution times are presently protracted. Moreover, we underscore that augmenting quantum computational capacity and the magnitude of parallelism can markedly ameliorate the performance of quantum machine learning algorithms. This inquiry furnishes invaluable insights regarding the extant scenario and future potentiality of machine learning applications in the quantum epoch. Colab: https://t.ly/QKuz0
Quantum machine learning has emerged as a promising frontier, leveraging the principles of quantum computing to enhance classical algorithms. This paper proposes an optimized Quantum Support Vector Classifier (QSVC) for the classification of the MNIST dataset (a benchmark in handwritten digit recognition). We employ a Chebyshev Parameterized Quantum Circuit (PQC) with an increased number of qubits and layers, and utilize an existing projected quantum kernel method to effectively capture the complex patterns in the data. To enhance model robustness, data augmentation techniques are applied, and dimensionality reduction is performed using Principal Component Analysis (PCA). The proposed QSVC demonstrates competitive accuracy on both the training and testing datasets, highlighting the potential of quantum kernel methods in practical machine learning applications. A comparative analysis with classical Support Vector Machines (SVM) demonstrates the feasibility of quantum methods for high-dimensional classification tasks. Additionally, we investigate the impact of varying the number of qubits in the quantum circuit, providing in-depth insights into the scalability and performance trade-offs of quantum kernel methods.