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

arXiv:2608.26224

Classical and Hybrid Quantum Machine Learning for Trigger-Like Event Selection on CMS Open Data: An Eight-Qubit, PCA-Constrained Benchmark

Tariq Mahmood, Muhammad Awais Rafique, T. Hussain, Juan Pablo Pérez Aguilar, Alfredo Raya, Muhammad Ahsan

Published Aug 26, 2026• 0 citations
arXivHTMLPDF

Abstract

Event triggering sits at the heart of high-energy physics, where the rare events of interest must be retained while an overwhelming background is discarded under tight latency and bandwidth budgets. This work compares four classical machine learning models, namely a support vector machine, an artificial neural network, a convolutional network and a long short-term memory network, with four hybrid quantum counterparts, on a trigger-like binary classification task built from CMS open data. The label is defined by an invariant-mass window, and the inputs combine reconstructed kinematics with physics-motivated derived variables: the pseudorapidity difference, the wrapped azimuthal difference, the angular separation and the total transverse momentum. The quantum models run under a fixed resource budget of eight qubits, a principal-component compression to sixteen features and state-vector simulation. Every model shares the same stratified split, the same preprocessing and a common decision threshold, and performance is reported through accuracy, ROC-AUC, F1-score, precision and recall. The strongest classical model is the artificial neural network, at 93.53 percent accuracy and 0.9819 ROC-AUC, while the strongest quantum model is the quantum convolutional network, at 90.89 percent accuracy and 0.9731 ROC-AUC, with the quantum neural network close behind. The quantum-kernel and recurrent quantum approaches trail both, which places the trainable hybrid embeddings ahead within this budget. The study is meant as a controlled reference point rather than a claim of quantum advantage.

DOI
10.48550/arxiv.2608.26224

Related papers

Comparing Classical and Quantum Machine Learning for Regression in High Energy Physics Collision Data

Tariq Mahmood, Zain ul Abidin, Itzel Luviano Soto +1 more

The classification and regression of particle collision events constitute a persistent computational challenge in experimental high energy physics, where large volumes of simulated data must be processed with both speed and precision. This work carries out a systematic comparison of four classical machine learning architectures, support vector machines (SVM), artificial neural networks (ANN), convolutional neural networks (CNN), and long short-term memory (LSTM) networks against their quantum counterparts: quantum SVM (QSVM), quantum neural networks (QNN), quantum CNN (QCNN), and quantum LSTM (QLSTM). All models are trained on simulated proton-proton collision events with electron-positron and muon-antimuon final states from the CERN Open Data portal, using transverse-momentum components as input features and transverse-momentum magnitude as the regression target. Classical architectures, and in particular the CNN and LSTM, achieve marginally better quantitative performance under current hardware and dataset constraints. Quantum models, however, reach competitive accuracy with substantially fewer trainable parameters: the QCNN reproduces the performance of the deep classical CNN using only four qubits and a circuit of depth three, pointing to a genuine parameter-efficiency advantage on near-term quantum devices. A baseline analysis confirms that the regression problem is non-trivial for shallow polynomial fits, supporting the relevance of the architectural comparison. These results characterize the trade-offs between classical and quantum approaches under realistic, resource-constrained conditions and provide a benchmark for future studies on actual quantum hardware.

cs.LGCross-listedarXiv:2608.28084•Aug 31, 2026•0 citations

Hybrid Quantum-Classical Machine Learning For Scalable AI Applications

Manish Patil, Shraddha Sanjay Mhetre

Background: The emergent need for scalable artificial intelligence (AI) models emphasizes the shortcomings of traditional machine learning in dealing with big, complicated datasets owing to computation and energy limitations. Objective: The architecture takes advantage of quantum encoding methods like amplitude and angle encoding for compact feature representation, with classical preprocessing providing compatibility and stability. Methods: To overcome these issues, this research introduces a Hybrid Quantum-Classical Machine Learning (HQML) framework that combines Quantum Variational Classifiers (QVCs) and Quantum Convolutional Neural Networks (QCNNs) with classical deep learning layers. Results: Experiments were performed on pre-curated datasets taken from the Scikit Machine Learning Repository, with hybrid models compared to classical Convolutional Neural Networks (CNNs). The implementation was done with PennyLane using TensorFlow Quantum for compatibility with noisy intermediate-scale quantum (NISQ) machines. The comparative analysis identified that the suggested HQML model achieves significant performance gains, as classification accuracy increased from 73 percent in the baseline classical CNN to 98% in the hybrid model. Precision also went up from 77 to 96 percent, recall from 81 to 100 percent, and F1-score from 79 to 98 percent. These findings confirm the ability of the model to leverage quantum benefits for extracting complicated data distributions without losing stability through classical components. The findings confirm the viability of HQML to real-world finance, healthcare, and intelligent automation applications where scalability and precision matter. Conclusion: This research establishes that quantum-classical integration not only minimizes training overhead but also attains near-perfect classification efficiency with constrained qubit resources. The research aids in filling the gap between theoretical quantum potential and deployable practical AI, opening the door for next-generation quantum-boosted intelligence.

International Journal of Advances in Signal and Image Sciences•Sep 25, 2025•0 citations

Quantum Machine Learning for Image Classification: A Controlled Benchmark

Kuan Huang, Meng Xu, Yingfeng Wang

Image-based quantum machine learning usually follows a hybrid pipeline: a classical encoder compresses an image, and a quantum circuit then acts as a kernel, feature map, trainable classifier, or measurement-based readout. Yet it remains unclear which of these roles contributes useful predictive structure because prior studies often change the input representation, classical baseline, and measurement budget simultaneously. We present a controlled benchmark spanning five roles: quantum-kernel geometry, fixed product and entangled features, trainable data-reuploading representations, shot-aware probability readout, and task-adaptive Pauli measurement. Six quantum-derived methods, six matched classical controls, and two unmatched neural references are evaluated on MNIST, Fashion-MNIST, CIFAR-10, and BreastMNIST using a shared frozen ResNet-18 and train-only PCA interface. We assess exact and finite-shot inference, predictive performance, calibration, selective risk, resource requirements, and statistical reliability. PdrQC-Matched was the strongest quantum-derived method on all four datasets. C-RBF was better on Fashion-MNIST and CIFAR-10 by 2.19 and 3.50 accuracy points, while BreastMNIST showed no detectable difference. Despite lower exact accuracy, PdrQC-Matched achieved significantly lower Brier score and AURC than C-RBF on Fashion-MNIST and CIFAR-10. Finite-shot results showed no automatic benefit from entanglement or shot-aware training. Overall, image-QML conclusions depend on quantum role, measurement budget, classical control, and evaluation metric.

Proceedings of the AAAI Symposium Series•Nov 6, 2026•0 citations

Do We Really Need Quantum Machine Learning?: A Multidimensional Empirical Study

Sudip Vhaduri, Ryan Gammon, Sayanton Dibbo

The rapid growth of computer vision and increasingly complex image recognition tasks has exposed fundamental computational limitations of classical machine learning models, motivating the exploration of quantum computing as an emerging new paradigm. This paper presents a comprehensive benchmarking study of classical and quantum machine learning models for image recognition on the MNIST handwritten digit dataset, evaluating both traditional models, a Classical Support Vector Machine (CSVM) and a Quantum Support Vector Machine (QSVM), and deep neural network models, a Classical Convolutional Neural Network (CCNN) and a Quantum Convolutional Neural Network (QCNN), across four performance dimensions: classification accuracy, computational runtime, parameter count, and memory requirements. Experiments are conducted as functions of both feature dimensionality and sample size, and across CPU and GPU execution environments, providing a controlled, multidimensional comparison to address gaps in prior work. For the SVM-based models, QSVM consistently outperforms CSVM in accuracy, reaching $\sim$ 0.90 versus $\sim$ 0.85 at 1,000 samples, with a higher computational cost. A feature count of 10 qubits and a sample size in the range of 200 -- 500 emerge as practical operating points that balance accuracy and runtime. For the neural network models, CCNN and QCNN achieve comparable classification accuracy, both exceeding 0.96 at 64 features and 60,000 samples, yet QCNN offers superior parameter and memory efficiency at higher feature counts, while incurring higher runtime. Across both model families, quantum models consistently outperform classical models by greater margins in accuracy as feature dimensionality or sample size increases.

cs.CVCross-listedarXiv:2605.27923•May 28, 2026•4 citations

Comparative Study of Classical and Quantum Machine Learning Models: Insights into Quantum Advantage in Materials Informatics

Aris Tri Jaka Harjanta, Hindriyanto Dwi Purnomo, Hendry

Quantum Machine Learning (QML) has emerged as a promising paradigm for addressing increasing computational and representational demands in materials informatics. While classical models such as Support Vector Machines (SVM) achieve strong predictive performance, they often struggle to capture complex, highly correlated interactions in high-dimensional materials data. QML addresses this challenge by leveraging quantum-mechanical principles to construct expressive feature embeddings, where prospective quantum advantage lies in generating feature spaces that are difficult to approximate classically. In this study, 1,000 crystalline compounds from the Open Quantum Materials Database (OQMD) are evaluated in a binary classification task based on formation-energy stability. The dataset is normalized, reduced to four dimensions via Principal Component Analysis (PCA), and encoded into quantum circuits. Three QML models—QSVM, VQC, and QNN—are benchmarked against a classical SVM using repeated stratified evaluation. Results show that the classical SVM achieves the highest accuracy (91.8% ± 0.012), followed by QSVM (60.8% ± 0.035), while VQC and QNN perform significantly worse. This gap is driven by limited qubit capacity, encoding inefficiencies, restricted circuit expressivity, and optimization challenges. Nevertheless, QSVM demonstrates stable performance, suggesting that potential quantum advantage may emerge from improved feature encoding and kernel design rather than deeper variational circuits.

Advance Sustainable Science Engineering and Technology•Apr 30, 2026•0 citations

A Parameter-Matched Comparison of Quantum and Classical Neural Networks for Classifying Simulated Supersymmetric Events

Arin Raval

Supersymmetry (SUSY) is a proposed hypothetical extension of the Standard Model (SM) that relates bosons and fermions differing in spin by one-half. Certain SUSY models may help address the hierarchy problem, provide a dark-matter candidate, and achieve gauge-coupling unification. Quantum machine learning (QML) has been investigated as a possible tool for high-energy physics (HEP) event classification, including the discrimination of SUSY signal from SM background. Rather than assessing the general feasibility of variational quantum classifiers, this study performs a strict, parameter-matched comparison between a classical Artificial Neural Network (ANN) and a classically simulated Quantum Neural Network (QNN), also considering training time as a point of comparison. Both models were trained on the same 18 input features from a 5000-event subset of the SUSY benchmark dataset, split 70/15/15 into stratified training, validation and test sets. The compact classical ANN (18→3→1 architecture) contained 61 trainable parameters; the QNN---amplitude embedding on five qubits, six variational blocks of RY - RX rotations, and a trainable classical output layer---contained 66 trainable parameters. Each model was optimised using the Adam optimiser with a learning rate of 1*10^{-3} and binary cross-entropy loss for 200 epochs, repeated across 100 independent training runs, each containing a different split sample. Mean test metrics were highly comparable: the QNN achieved marginally higher accuracy (79.0% vs 78.9%), recall and F1 score, while the ANN achieved higher precision, higher AUC (0.858 vs. 0.857) and lower loss (0.455 vs. 0.459), with no metric differing by more than 0.8 percentage points. However, the ANN completed its 100 training runs in 20 minutes, versus around 5 hours for the classically simulated QNN, taking almost 15 times as long to perform the same task. These results do not imply a quantum advantage, but confirm that a compact variational QNN can learn to distinguish SUSY signal from SM background at a level comparable to that of a classical model, but at a computational cost that makes it far less efficient.

•Aug 25, 2026•0 citations
Event triggering sits at the heart of high-energy physics, where the rare events of interest must be retained while an overwhelming background is discarded under tight latency and bandwidth budgets. This work compares four classical machine learning models, namely a support vector machine, an artificial neural network, a convolutional network and a long short-term memory network, with four hybrid quantum counterparts, on a trigger-like binary classification task built from CMS open data. The label is defined by an invariant-mass window, and the inputs combine reconstructed kinematics with physics-motivated derived variables: the pseudorapidity difference, the wrapped azimuthal difference, the angular separation and the total transverse momentum. The quantum models run under a fixed resource budget of eight qubits, a principal-component compression to sixteen features and state-vector simulation. Every model shares the same stratified split, the same preprocessing and a common decision threshold, and performance is reported through accuracy, ROC-AUC, F1-score, precision and recall. The strongest classical model is the artificial neural network, at 93.53 percent accuracy and 0.9819 ROC-AUC, while the strongest quantum model is the quantum convolutional network, at 90.89 percent accuracy and 0.9731 ROC-AUC, with the quantum neural network close behind. The quantum-kernel and recurrent quantum approaches trail both, which places the trainable hybrid embeddings ahead within this budget. The study is meant as a controlled reference point rather than a claim of quantum advantage.
The classification and regression of particle collision events constitute a persistent computational challenge in experimental high energy physics, where large volumes of simulated data must be processed with both speed and precision. This work carries out a systematic comparison of four classical machine learning architectures, support vector machines (SVM), artificial neural networks (ANN), convolutional neural networks (CNN), and long short-term memory (LSTM) networks against their quantum counterparts: quantum SVM (QSVM), quantum neural networks (QNN), quantum CNN (QCNN), and quantum LSTM (QLSTM). All models are trained on simulated proton-proton collision events with electron-positron and muon-antimuon final states from the CERN Open Data portal, using transverse-momentum components as input features and transverse-momentum magnitude as the regression target. Classical architectures, and in particular the CNN and LSTM, achieve marginally better quantitative performance under current hardware and dataset constraints. Quantum models, however, reach competitive accuracy with substantially fewer trainable parameters: the QCNN reproduces the performance of the deep classical CNN using only four qubits and a circuit of depth three, pointing to a genuine parameter-efficiency advantage on near-term quantum devices. A baseline analysis confirms that the regression problem is non-trivial for shallow polynomial fits, supporting the relevance of the architectural comparison. These results characterize the trade-offs between classical and quantum approaches under realistic, resource-constrained conditions and provide a benchmark for future studies on actual quantum hardware.
Background: The emergent need for scalable artificial intelligence (AI) models emphasizes the shortcomings of traditional machine learning in dealing with big, complicated datasets owing to computation and energy limitations. Objective: The architecture takes advantage of quantum encoding methods like amplitude and angle encoding for compact feature representation, with classical preprocessing providing compatibility and stability. Methods: To overcome these issues, this research introduces a Hybrid Quantum-Classical Machine Learning (HQML) framework that combines Quantum Variational Classifiers (QVCs) and Quantum Convolutional Neural Networks (QCNNs) with classical deep learning layers. Results: Experiments were performed on pre-curated datasets taken from the Scikit Machine Learning Repository, with hybrid models compared to classical Convolutional Neural Networks (CNNs). The implementation was done with PennyLane using TensorFlow Quantum for compatibility with noisy intermediate-scale quantum (NISQ) machines. The comparative analysis identified that the suggested HQML model achieves significant performance gains, as classification accuracy increased from 73 percent in the baseline classical CNN to 98% in the hybrid model. Precision also went up from 77 to 96 percent, recall from 81 to 100 percent, and F1-score from 79 to 98 percent. These findings confirm the ability of the model to leverage quantum benefits for extracting complicated data distributions without losing stability through classical components. The findings confirm the viability of HQML to real-world finance, healthcare, and intelligent automation applications where scalability and precision matter. Conclusion: This research establishes that quantum-classical integration not only minimizes training overhead but also attains near-perfect classification efficiency with constrained qubit resources. The research aids in filling the gap between theoretical quantum potential and deployable practical AI, opening the door for next-generation quantum-boosted intelligence.
Image-based quantum machine learning usually follows a hybrid pipeline: a classical encoder compresses an image, and a quantum circuit then acts as a kernel, feature map, trainable classifier, or measurement-based readout. Yet it remains unclear which of these roles contributes useful predictive structure because prior studies often change the input representation, classical baseline, and measurement budget simultaneously. We present a controlled benchmark spanning five roles: quantum-kernel geometry, fixed product and entangled features, trainable data-reuploading representations, shot-aware probability readout, and task-adaptive Pauli measurement. Six quantum-derived methods, six matched classical controls, and two unmatched neural references are evaluated on MNIST, Fashion-MNIST, CIFAR-10, and BreastMNIST using a shared frozen ResNet-18 and train-only PCA interface. We assess exact and finite-shot inference, predictive performance, calibration, selective risk, resource requirements, and statistical reliability. PdrQC-Matched was the strongest quantum-derived method on all four datasets. C-RBF was better on Fashion-MNIST and CIFAR-10 by 2.19 and 3.50 accuracy points, while BreastMNIST showed no detectable difference. Despite lower exact accuracy, PdrQC-Matched achieved significantly lower Brier score and AURC than C-RBF on Fashion-MNIST and CIFAR-10. Finite-shot results showed no automatic benefit from entanglement or shot-aware training. Overall, image-QML conclusions depend on quantum role, measurement budget, classical control, and evaluation metric.
The rapid growth of computer vision and increasingly complex image recognition tasks has exposed fundamental computational limitations of classical machine learning models, motivating the exploration of quantum computing as an emerging new paradigm. This paper presents a comprehensive benchmarking study of classical and quantum machine learning models for image recognition on the MNIST handwritten digit dataset, evaluating both traditional models, a Classical Support Vector Machine (CSVM) and a Quantum Support Vector Machine (QSVM), and deep neural network models, a Classical Convolutional Neural Network (CCNN) and a Quantum Convolutional Neural Network (QCNN), across four performance dimensions: classification accuracy, computational runtime, parameter count, and memory requirements. Experiments are conducted as functions of both feature dimensionality and sample size, and across CPU and GPU execution environments, providing a controlled, multidimensional comparison to address gaps in prior work. For the SVM-based models, QSVM consistently outperforms CSVM in accuracy, reaching ∼\sim∼ 0.90 versus ∼\sim∼ 0.85 at 1,000 samples, with a higher computational cost. A feature count of 10 qubits and a sample size in the range of 200 -- 500 emerge as practical operating points that balance accuracy and runtime. For the neural network models, CCNN and QCNN achieve comparable classification accuracy, both exceeding 0.96 at 64 features and 60,000 samples, yet QCNN offers superior parameter and memory efficiency at higher feature counts, while incurring higher runtime. Across both model families, quantum models consistently outperform classical models by greater margins in accuracy as feature dimensionality or sample size increases.
Quantum Machine Learning (QML) has emerged as a promising paradigm for addressing increasing computational and representational demands in materials informatics. While classical models such as Support Vector Machines (SVM) achieve strong predictive performance, they often struggle to capture complex, highly correlated interactions in high-dimensional materials data. QML addresses this challenge by leveraging quantum-mechanical principles to construct expressive feature embeddings, where prospective quantum advantage lies in generating feature spaces that are difficult to approximate classically. In this study, 1,000 crystalline compounds from the Open Quantum Materials Database (OQMD) are evaluated in a binary classification task based on formation-energy stability. The dataset is normalized, reduced to four dimensions via Principal Component Analysis (PCA), and encoded into quantum circuits. Three QML models—QSVM, VQC, and QNN—are benchmarked against a classical SVM using repeated stratified evaluation. Results show that the classical SVM achieves the highest accuracy (91.8% ± 0.012), followed by QSVM (60.8% ± 0.035), while VQC and QNN perform significantly worse. This gap is driven by limited qubit capacity, encoding inefficiencies, restricted circuit expressivity, and optimization challenges. Nevertheless, QSVM demonstrates stable performance, suggesting that potential quantum advantage may emerge from improved feature encoding and kernel design rather than deeper variational circuits.
Supersymmetry (SUSY) is a proposed hypothetical extension of the Standard Model (SM) that relates bosons and fermions differing in spin by one-half. Certain SUSY models may help address the hierarchy problem, provide a dark-matter candidate, and achieve gauge-coupling unification. Quantum machine learning (QML) has been investigated as a possible tool for high-energy physics (HEP) event classification, including the discrimination of SUSY signal from SM background. Rather than assessing the general feasibility of variational quantum classifiers, this study performs a strict, parameter-matched comparison between a classical Artificial Neural Network (ANN) and a classically simulated Quantum Neural Network (QNN), also considering training time as a point of comparison. Both models were trained on the same 18 input features from a 5000-event subset of the SUSY benchmark dataset, split 70/15/15 into stratified training, validation and test sets. The compact classical ANN (18→3→1 architecture) contained 61 trainable parameters; the QNN---amplitude embedding on five qubits, six variational blocks of RY - RX rotations, and a trainable classical output layer---contained 66 trainable parameters. Each model was optimised using the Adam optimiser with a learning rate of 1*10^{-3} and binary cross-entropy loss for 200 epochs, repeated across 100 independent training runs, each containing a different split sample. Mean test metrics were highly comparable: the QNN achieved marginally higher accuracy (79.0% vs 78.9%), recall and F1 score, while the ANN achieved higher precision, higher AUC (0.858 vs. 0.857) and lower loss (0.455 vs. 0.459), with no metric differing by more than 0.8 percentage points. However, the ANN completed its 100 training runs in 20 minutes, versus around 5 hours for the classically simulated QNN, taking almost 15 times as long to perform the same task. These results do not imply a quantum advantage, but confirm that a compact variational QNN can learn to distinguish SUSY signal from SM background at a level comparable to that of a classical model, but at a computational cost that makes it far less efficient.