Quantum classification faces two key challenges. First, the difficulty of distinguishing between different classes varies: some class pairs are easy to separate, while others are more challenging. Second, practical execution is affected by noise, finite sampling, and measurement overhead. To address these issues, we propose PAPUS, a framework for pair-adaptive quantum classification in Pauli space. The method evaluates candidate upload circuits using low-weight Pauli features and formulates upload design as a structured model selection problem based on discriminative representations. By dynamically adjusting circuit complexity according to class-pair difficulty, the framework achieves a better balance between classification accuracy and resource efficiency. Experiments on 9 data sets with 474 tasks show that PAPUS achieves a favorable balance between predictive performance and execution cost. Specifically, PAPUS attains classification accuracies above 90% in both local noiseless simulation and the IonQ noisy simulator, while requiring substantially lower measurement and circuit cost (fewer total measurement shots and fewer quantum gates for data upload). Compared with the two conventional baselines, template_cv and kta_exact, PAPUS also shows much stronger robustness under noise: accuracy decreases by only 1.67% in the noisy setting, whereas both baselines degrade by 9.44%.
Quantum classification faces two key challenges. First, the difficulty of distinguishing between different classes varies: some class pairs are easy to separate, while others are more challenging. Second, practical execution is affected by noise, finite sampling, and measurement overhead. To address these issues, we propose PAPUS, a framework for pair-adaptive quantum classification in Pauli space. The method evaluates candidate upload circuits using low-weight Pauli features and formulates upload design as a structured model selection problem based on discriminative representations. By dynamically adjusting circuit complexity according to class-pair difficulty, the framework achieves a better balance between classification accuracy and resource efficiency. Experiments on 9 data sets with 474 tasks show that PAPUS achieves a favorable balance between predictive performance and execution cost. Specifically, PAPUS attains classification accuracies above 90% in both local noiseless simulation and the IonQ noisy simulator, while requiring substantially lower measurement and circuit cost (fewer total measurement shots and fewer quantum gates for data upload). Compared with the two conventional baselines, template_cv and kta_exact, PAPUS also shows much stronger robustness under noise: accuracy decreases by only 1.67% in the noisy setting, whereas both baselines degrade by 9.44%.
Abstract Hybrid Quantum-Classical Convolutional Neural Networks (HQCNNs) show significant potential for image classification by embedding quantum circuits as convolutional kernels within classical CNNs, leveraging quantum computing to enhance feature extraction. While this approach provides a feasible pathway toward achieving quantum advantage under current Noisy Intermediate-Scale Quantum (NISQ) hardware constraints, existing HQCNNs face critical challenges, including limited classification accuracy and excessive quantum resource consumption when processing multi-channel data. To address these issues, this paper proposes the Quantum Multi-channel Data Uploading Convolution (QMDUC) method, which employs data re-uploading techniques to optimize quantum data encoding strategies, significantly reducing qubit requirements while maintaining complex data processing capabilities. QMDUC supports two encoding modes: Channel-Priority QMDUC (CP-QMDUC) and Data-Priority QMDUC (DP-QMDUC), enabling flexible trade-offs between qubit count and circuit depth, thereby providing adaptive solutions tailored to different quantum hardware constraints. Experimental validation on the CIFAR-10 three-channel color image dataset and Fashion-MNIST single-channel grayscale image dataset demonstrates that QMDUC significantly outperforms existing HQCNN methods in both classification accuracy and quantum resource efficiency. Specifically, under low-resource configurations, the proposed method achieves an approximate reduction in qubit requirements of 75%–95% while maintaining comparable or superior accuracy; under higher-resource configurations, it achieves approximately 10% improvement in peak accuracy with comparable qubit consumption to existing methods, providing an effective technical pathway for practical applications of quantum machine learning in computer vision.
As a generalized quantum machine learning model, parameterized quantum circuits (PQC) have been found to perform poorly in terms of classification accuracy and model scalability for multi-category classification tasks. To address this issue, we propose a scalable parameterized quantum circuits classifier (SPQCC), which performs per-channel PQC and combines the measurements as the output of the trainable parameters of the classifier. By minimizing the cross-entropy loss through optimizing the trainable parameters of PQC, SPQCC leads to a fast convergence of the classifier. The parallel execution of identical PQCs on different quantum machines with the same structure and scale reduces the complexity of classifier design. Classification simulations performed on the MNIST Dataset show that the accuracy of our proposed classifier far exceeds that of other quantum classification algorithms, achieving the state-of-the-art simulation result and surpassing/reaching classical classifiers with a considerable number of trainable parameters. Our classifier demonstrates excellent scalability and classification performance.
Abstract Hybrid quantum machine learning combines classical minimization with parameterized quantum circuits, enabling powerful learning models even on near-term quantum hardware. This work presents an evaluation of hybrid quantum machine learning classifiers using the data re-uploading framework across nine binary and multi-class classification tasks with varying geometric complexity. We compare three quantum circuit architectures, single-qubit, two-qubit, and two-qubit entangled, trained with three classical minimizers: COBYLA, Nelder-Mead, and SLSQP. Our experiments span training sizes from 1 to 200 samples and reveal that two-qubit classifiers, when paired with the SLSQP minimizer, achieve the highest overall test accuracy of 99.775% on structured tasks and strong performance (up to 89.78%) on complex multi-class pattern. Among minimizers, SLSQP consistently provides the best generalization, while entanglement improves accuracy in non-linear and fragmented tasks. By analyzing accuracy trends, generalization gaps, and confusion matrices, this study offers practical guidance for designing quantum classifiers that scale effectively and generalize well, providing a foundation for real-world Hybrid quantum machine learning applications.
Neural quantum kernels (NQKs) construct quantum kernels by pretraining a quantum neural network (QNN) and subsequently reusing the trained circuit as a task-adapted embedding. Extending this framework to qudits, with local unitaries in SU(d), provides a natural route to richer data embeddings through the increased local degrees of freedom and a direct interface for multiclass classification via intrinsically multi-level quantum systems. In this work, focusing on qutrits (d=3), we extend NQKs to the qudit setting and perform a systematic study of key design choices, including the number of encoded features, the number of qutrits, the kernel construction (1-to-n and n-to-n), and the parameterization of SU(3) unitaries. Across binary and three-class tasks on four benchmark datasets, qutrit NQKs improve over the corresponding QNN baselines in nearly all settings considered and can benefit from scaling both the feature budget and the system size, although the magnitude of these gains may saturate, is dataset-dependent, and depends on the chosen parameterization. In particular, an ablation over SU(3) parameterizations shows that the unitary representation can substantially impact both optimization behaviour and classifier performance. These findings highlight the potential of qudit-based quantum models not only as a straightforward generalization of qubit-based architectures, but also as a promising means to better exploit complex data structures in quantum machine learning.
Abstract In the field of machine learning, the multi-category classification problem plays a crucial role. Solving the problem has a profound impact on driving the innovation and development of machine learning techniques and addressing complex problems in the real world. In recent years, researchers have begun to focus on utilizing quantum computing to solve the multi-category classification problem. Some studies have shown that the process of processing information in the brain may be related to quantum phenomena, with different brain regions having neurons with different structures. Inspired by this, we design a quantum multi-category classifier model from this perspective for the first time. The model employs a heterogeneous population of quantum neural networks (QNNs) to simulate the cooperative work of multiple different brain regions. When processing information, these heterogeneous clusters of QNNs allow for simultaneous execution on different quantum computers, thus simulating the brain’s ability to utilize multiple brain regions working in concert to maintain the robustness of the model. By setting the number of heterogeneous QNN clusters and parameterizing the number of stacks of unit layers in the quantum circuit, the model demonstrates excellent scalability in dealing with different types of data and different numbers of classes in the classification problem. Based on the attention mechanism of the brain, we integrate the processing results of heterogeneous QNN clusters to achieve high accuracy in classification. Finally, we conducted classification simulation experiments on different datasets. The results show that our method exhibits strong robustness and scalability. Among them, on different subsets of the MNIST dataset, its classification accuracy improves by up to about 5% compared to other quantum multiclassification algorithms. This result becomes the state-of-the-art simulation result for quantum classification models and exceeds the performance of classical classifiers with a considerable number of trainable parameters on some subsets of the MNIST dataset.
Abstract Image classification, a pivotal task in multiple industries, faces computational challenges due to the burgeoning volume of visual data. This research addresses these challenges by introducing two quantum machine learning models that leverage the principles of quantum mechanics for effective computations. Our first model, a hybrid quantum neural network with parallel quantum circuits, enables the execution of computations even in the noisy intermediate-scale quantum era, where circuits with a large number of qubits are currently infeasible. This model demonstrated a record-breaking classification accuracy of 99.21% on the full MNIST dataset, surpassing the performance of known quantum–classical models, while having eight times fewer parameters than its classical counterpart. Also, the results of testing this hybrid model on a Medical MNIST (classification accuracy over 99%), and on CIFAR-10 (classification accuracy over 82%), can serve as evidence of the generalizability of the model and highlights the efficiency of quantum layers in distinguishing common features of input data. Our second model introduces a hybrid quantum neural network with a Quanvolutional layer, reducing image resolution via a convolution process. The model matches the performance of its classical counterpart, having four times fewer trainable parameters, and outperforms a classical model with equal weight parameters. These models represent advancements in quantum machine learning research and illuminate the path towards more accurate image classification systems.