Quantum machine learning algorithms hold promise for near-term quantum advantage, yet theirperformance under realistic hardware noise remains poorly characterized. In this work, we introduce ahybrid classical-to-quantum transfer learning architecture in which a fixed linear encoder maps twodimensional classical input data to four quantum rotation angles, which are subsequently encoded into aparameterized quantum circuit (PQC) ansatz executed on the PKTRON v3.7.3 simulation frameworkusing the PK NoisyLab 8Q virtual device. We systematically evaluate the degradation of binaryclassification accuracy and the quantum feature space under three distinct decoherence channels —amplitude damping, phase damping, and depolarizing noise — across five noise strength levels (p = 0.01to 0.35) and three entanglement depths (L = 1, 2, 3 layers). Our results reveal that phase damping anddepolarizing noise impose an immediate, strength-invariant accuracy reduction of 10% from the idealbaseline of 83.33%, while amplitude damping exhibits partial robustness at low noise levels. Frobeniusdistance analysis demonstrates that depolarizing noise induces the steepest degradation of the quantumfeature space, reaching a saturation plateau at p = 0.20. Critically, deeper entanglement circuits achievehigher ideal accuracy but exhibit superlinear growth in Frobenius distance under noise, suggesting afundamental accuracy-robustness tradeoff. We propose a Noise Sensitivity Score S(N) as a lightweight,application-level noise diagnostic metric, offering a tractable alternative to full quantum processtomography for NISQ-era devices.
Quantum machine learning algorithms hold promise for near-term quantum advantage, yet theirperformance under realistic hardware noise remains poorly characterized. In this work, we introduce ahybrid classical-to-quantum transfer learning architecture in which a fixed linear encoder maps twodimensional classical input data to four quantum rotation angles, which are subsequently encoded into aparameterized quantum circuit (PQC) ansatz executed on the PKTRON v3.7.3 simulation frameworkusing the PK NoisyLab 8Q virtual device. We systematically evaluate the degradation of binaryclassification accuracy and the quantum feature space under three distinct decoherence channels —amplitude damping, phase damping, and depolarizing noise — across five noise strength levels (p = 0.01to 0.35) and three entanglement depths (L = 1, 2, 3 layers). Our results reveal that phase damping anddepolarizing noise impose an immediate, strength-invariant accuracy reduction of 10% from the idealbaseline of 83.33%, while amplitude damping exhibits partial robustness at low noise levels. Frobeniusdistance analysis demonstrates that depolarizing noise induces the steepest degradation of the quantumfeature space, reaching a saturation plateau at p = 0.20. Critically, deeper entanglement circuits achievehigher ideal accuracy but exhibit superlinear growth in Frobenius distance under noise, suggesting afundamental accuracy-robustness tradeoff. We propose a Noise Sensitivity Score S(N) as a lightweight,application-level noise diagnostic metric, offering a tractable alternative to full quantum processtomography for NISQ-era devices.
Abstract Quantum machine learning algorithms hold promise for near-term quantum advantage, yet their performance under realistic hardware noise remains poorly characterized. In this work, we introduce a hybrid classical-to-quantum transfer learning architecture in which a fixed linear encoder maps twodimensional classical input data to four quantum rotation angles, which are subsequently encoded into a parameterized quantum circuit (PQC) ansatz executed on the PKTRON v3.7.3 simulation framework using the PK NoisyLab 8Q virtual device. We systematically evaluate the degradation of binary classification accuracy and the quantum feature space under three distinct decoherence channels — amplitude damping, phase damping, and depolarizing noise — across five noise strength levels (p = 0.01 to 0.35) and three entanglement depths (L = 1, 2, 3 layers). Our results reveal that phase damping and depolarizing noise impose an immediate, strength-invariant accuracy reduction of 10% from the ideal baseline of 83.33%, while amplitude damping exhibits partial robustness at low noise levels. Frobenius distance analysis demonstrates that depolarizing noise induces the steepest degradation of the quantum feature space, reaching a saturation plateau at p = 0.20. Critically, deeper entanglement circuits achieve higher ideal accuracy but exhibit superlinear growth in Frobenius distance under noise, suggesting a fundamental accuracy-robustness tradeoff. We propose a Noise Sensitivity Score S(N) as a lightweight, application-level noise diagnostic metric, offering a tractable alternative to full quantum process tomography for NISQ-era devices.
Near-term quantum machine learning (QML) models operate in environments wherein noise is unavoidable, arising from both imperfect classical data acquisition and the limitations of noisy intermediate-scale quantum (NISQ) hardware. Although most existing studies have focused primarily on quantum circuit noise in isolation, the combined influence of corrupted classical inputs and quantum hardware noise has received comparatively little attention.
In this work, we present a systematic experimental study of the robustness of a variational quantum classifier under realistic multi-level noise conditions. Using the Titanic dataset as a benchmark, a range of dataset-level noise models-including speckle noise, impulse noise, quantization noise, and feature dropout are applied to classical features prior to quantum encoding using a ZZ feature map. In parallel, hardware-inspired quantum noise channels such as depolarizing noise, amplitude damping, phase damping, Pauli errors, and readout errors are incorporated at the circuit level using the Qiskit Aer simulator.
The experimental results indicate that noise in classical input data can significantly intensify the effects of quantum decoherence, resulting in less stable training and noticeably lower classification accuracy. Together, these observations emphasize the importance of designing and evaluating quantum machine learning pipelines with noise in mind, and highlight the need to consider classical and quantum noise simultaneously when assessing QML performance in the NISQ era
In current noisy intermediate-scale quantum (NISQ) devices, hybrid quantum neural networks (HQNNs) offer a promising solution, combining the strengths of classical machine learning with quantum computing capabilities. However, the performance of these networks can be significantly affected by the quantum noise inherent in NISQ devices. In this paper, we conduct an extensive comparative analysis of various HQNN algorithms, namely Quantum Convolution Neural Network (QCNN), Quanvolutional Neural Network (QuanNN), and Quantum Transfer Learning (QTL), for image classification tasks. We evaluate the performance of each algorithm across quantum circuits with different entangling structures, variations in layer count, and optimal placement in the architecture. Subsequently, we select the highest-performing architectures and assess their robustness against noise influence by introducing quantum gate noise through Phase Flip, Bit Flip, Phase Damping, Amplitude Damping, and the Depolarization Channel. Our results reveal that the top-performing models exhibit varying resilience to different noise channels. However, in most scenarios, the QuanNN demonstrates greater robustness across various quantum noise channels, consistently outperforming other models. This highlights the importance of tailoring model selection to specific noise environments in NISQ devices.
Hybrid quantum-classical computing offers a promising pathway to harness quantum advantages while mitigating the hardware limitations of noisy intermediate-scale quantum (NISQ) devices. This work proposes a modular hybrid architecture that combines classical preprocessing, quantum-enhanced feature mapping through parameterized quantum circuits (PQCs), and adaptive classical optimization. The framework is designed to improve robustness and generalization of machine learning models in noisy environments, addressing critical issues such as overfitting, noise sensitivity, and hardware-induced errors. Experimental evaluations were conducted on benchmark datasets, including Iris, reduced-dimension MNIST, synthetic noisy XOR, Wine Quality, and Breast Cancer Wisconsin. The proposed model was tested under multiple noise models, such as Gaussian, salt-and-pepper, Poisson, and quantum-specific errors (bit-flip, phase-flip, and depolarizing), using 5-fold cross-validation and repeated trials for statistical rigor. Key performance metrics included Classification Accuracy (CA), Robustness Index (RI), and Training Stability (TS), with results reported as mean ± standard deviation and supported by paired t-tests (p < 0.05). Results demonstrate that the hybrid model consistently outperforms classical support vector machines (SVM) and standalone variational quantum classifiers (VQC) in noisy conditions, achieving smaller accuracy degradation and higher stability. The combination of quantum- enhanced feature mapping, optimizer synergy, and modular design is shown to be central to its resilience.
Gate-level noise degrades the classification accuracy of Quantum Support Vector Machines (QSVMs) on Noisy Intermediate-Scale Quantum (NISQ) hardware, and the degree of degradation depends on how classical data is encoded into quantum states. We tested Z, ZZ, a Pauli, and an amplitude-inspired feature maps under depolarizing, bit-flip, and phase-flip noise channels in 52 controlled experiments with error probabilities p=0.01,0.05,0.10, and 0.50. The amplitude-inspired feature map had 100\% test accuracy up to p=0.10 across all three noise channels, while other feature maps fell to 65-90\% under the same noise level and type. The Z feature map was found to be immune to phase-flip noise to a significantly high error rate, a consequence of the commutation relation [RZ,Z]=0. Entangled circuit variants produced generalization gaps in train-test sets of up to 17.5\% under noise, whereas the amplitude variants maintained zero gap throughout. These results give practitioners data-driven criteria for a feature map on near-term quantum hardware.
Quantum Neural Networks (QNNs) offer a compelling approach to quantum machine learning, but their effectiveness is heavily constrained by noise on Noisy Intermediate-Scale Quantum (NISQ) hardware. Prior studies examine global noise effects, yet the sensitivity of individual QNN layers remains underexplored.This paper presents a layer-wise noise sensitivity analysis framework that isolates the impact of depolarizing, amplitude damping, and phase damping noise on the encoding, variational, and measurement layers of QNNs. Using Qiskit Aer, we simulate targeted noise injection and evaluate performance on the Iris dataset. We introduce the Layerwise Sensitivity Drop (LSD) metric to quantify degradation under noise.Results show that the encoding layer is most vulnerable (LSD up to 42.9%), while the measurement layer is comparatively robust. Additionally, entangling gates are found to be more noisesensitive than rotational gates. These insights inform future QNN design strategies, emphasizing layer-specific error mitigation to improve resilience on near-term quantum hardware.