Quantum Machine Learning (QML) promises computational advantages over Deep Learning (DL), but its effectiveness in practical network Traffic Classification (TC) remains underexplored. In this paper, we provide the first systematic use of eXplainable Artificial Intelligence (XAI) to interpret QMLbased classifiers for (mobile) network TC. We combine (a) SHAPbased attribution and (b) calibration analysis to compare QML and DL models across both small- and large-scale TC tasks. This dual perspective enables a deeper understanding of QML interpretability and reliability relative to DL in network TC, exposing key limitations and avenues for improvement.
Quantum Machine Learning (QML) promises computational advantages over Deep Learning (DL), but its effectiveness in practical network Traffic Classification (TC) remains underexplored. In this paper, we provide the first systematic use of eXplainable Artificial Intelligence (XAI) to interpret QMLbased classifiers for (mobile) network TC. We combine (a) SHAPbased attribution and (b) calibration analysis to compare QML and DL models across both small- and large-scale TC tasks. This dual perspective enables a deeper understanding of QML interpretability and reliability relative to DL in network TC, exposing key limitations and avenues for improvement.
Network traffic analysis is essential for modern communication systems, focusing on tasks like traffic classification, prediction, and anomaly detection. While classical Machine Learning (ML) and Deep Learning (DL) methods have proven effective, their scalability and real-time performance can be limited by evolving traffic patterns and computational demands. Quantum Machine-Learning (QML) offers a promising alternative by utilizing quantum computing's parallelism. This paper examines QML's application in mobile traffic classification, comparing classical methods such as Multi-layer Perceptron (MLP) and Convolutional Neural Networks (CNNs) with Quantum Neural Networks (QNNs) using different embedding types. Our experiments, conducted on the MIRAGE-COVID-CCMA-2022 dataset, show that QNNs achieve competitive performance, indicating QML's potential for efficient large-scale traffic classification in future networks.
Abstract As machine learning grows increasingly complex due to big data and deep learning, model explainability has become essential to fostering user trust. Quantum machine learning (QML) has emerged as a promising field, leveraging quantum computing to enhance classical machine learning methods, particularly through quantum representation learning (QRL). QRL aims to provide more efficient and powerful machine learning capabilities on noisy intermediate-scale quantum (NISQ) devices. However, interpreting QRL models poses significant challenges due to the reliance on quantum gate-based parameterized circuits, which, while analogous to classical neural network layers, operate in the quantum domain. To address these challenges, we propose an explainable QRL framework combining a quantum autoencoder (QAE) with a variational quantum classifier (VQC) and incorporating theoretical and empirical explainability for image data. Our dual approach enhances model interpretability by integrating visual explanations via local interpretable model-agnostic explanations (LIME) and analytical insights using Shapley Additive Explanations (SHAP). These complementary methods provide a deeper understanding of the model’s decision-making process based on prediction outcomes. Experimental evaluations on simulators and superconducting quantum hardware validate the effectiveness of the proposed framework for classification tasks, underscoring the importance of explainable representation learning in advancing QML towards more transparent and reliable applications.
Abstract Methods of artificial intelligence (AI) and especially machine learning (ML) have been growing ever more complex, and at the same time have more and more impact on people’s lives. This leads to explainable AI (XAI) manifesting itself as an important research field that helps humans to better comprehend ML systems. In parallel, quantum machine learning (QML) is emerging with the ongoing improvement of quantum computing hardware combined with its increasing availability via cloud services. QML enables quantum-enhanced ML in which quantum mechanics is exploited to facilitate ML tasks, typically in the form of quantum-classical hybrid algorithms that combine quantum and classical resources. Quantum gates constitute the building blocks of gate-based quantum hardware and form circuits that can be used for quantum computations. For QML applications, quantum circuits are typically parameterized and their parameters are optimized classically such that a suitably defined objective function is minimized. Inspired by XAI, we raise the question of the explainability of such circuits by quantifying the importance of (groups of) gates for specific goals. To this end, we apply the well-established concept of Shapley values. The resulting attributions can be interpreted as explanations for why a specific circuit works well for a given task, improving the understanding of how to construct parameterized (or variational) quantum circuits, and fostering their human interpretability in general. An experimental evaluation on simulators and two superconducting quantum hardware devices demonstrates the benefits of the proposed framework for classification, generative modeling, transpilation, and optimization. Furthermore, our results shed some light on the role of specific gates in popular QML approaches.
Explainable artificial intelligence (XAI) remains underexplored in quantum machine learning (QML) due to limited quantum computational resources and the lack of practical interpretability techniques. In this study, we implement the Local Interpretable Model-Agnostic Explanations (LIME) algorithm within the CUDA-Q framework and propose an interpretable QML analysis method using GPU-accelerated quantum simulation. All experiments are conducted on a GPU-based simulator rather than real quantum hardware, enabling stable large-scale sampling required by LIME and ensuring reproducible results. We design a binary classification model based on a single-qubit variational circuit composed of parameterized RY and RX rotations, and the model is trained with gradient-based optimization. The performance is evaluated on the MNIST, KMNIST, Fashion-MNIST, and CIFAR-10 datasets using training and test accuracy as metrics. By applying LIME, the classification outcomes of the quantum model are visualized and interpreted in a human-understandable form, demonstrating that meaningful interpretability can be achieved even under limited qubit counts and constrained computational environments. Furthermore, we analyze the trade-off between simulator performance and model accuracy, confirming that GPU-based CUDA-Q simulation provides a scalable and practical research environment for QML. This work offers a concrete pathway toward transparent and interpretable quantum machine learning.
The rapid growth of ultra-large and highdimensional datasets has intensified the demand for deep learning (DL) models that are scalable, accurate, transparent, and interpretable. In parallel, quantum-inspired computing has emerged as a promising paradigm that emulates quantum principles such as superposition, entanglement, and probabilistic state evolution without requiring quantum hardware. This study presents a comprehensive survey of explainable quantum-inspired DL models for large-scale data mining. Existing quantum-inspired architectures integrated with deep neural networks (DNN), including convolutional, recurrent, and attention-based models, are systematically analyzed alongside explainable artificial intelligence (XAI) techniques to enhance interpretability and trustworthiness. The reviewed methods are classified based on learning paradigms, explainability mechanisms, scalability strategies, and application domains such as healthcare analytics, cybersecurity, intelligent transportation, and industrial big data systems. Furthermore, evaluation metrics, computational complexity, and performance trade-offs are examined in comparison with conventional DL approaches. Finally, key open challenges and future research directions for real-world large-scale deployment are highlighted.
This research paper explores the explainability of decision-making in hybrid quantum–classical classification models. Quantum Machine Learning (QML) combines the power of quantum computing with classical machine learning techniques to solve complex computational problems. However, understanding how these hybrid models make decisions remains a significant challenge. The study focuses on analyzing hybrid quantum–classical architectures and evaluating methods that improve the interpretability of their predictions. Techniques from Explainable Artificial Intelligence (XAI) are applied to investigate the internal decision-making processes of the models. The research also compares classical machine learning approaches with quantum-enhanced models to understand their performance and transparency. The results demonstrate that explainability techniques can help researchers better interpret model predictions, identify important features, and improve trust in hybrid quantum–classical systems. This work contributes to the development of transparent and reliable quantum machine learning applications in areas such as healthcare, finance, and data analytics.