This is a planned study that compares how much does data encoding techniques affect Quantum Machine Learning (QML) performance compared to Classical Machine Learning (CML)? Unlike classical models, QML cannot directly read a dataset; it must first convert features into quantum states using conversion methods like Classical machine learning reads normal numbers directly, but Quantum machine learning first turns those numbers into quantum form (qubits) that can be present in multiple states at once. This strengthens core AI skills for data preprocessing, PCA (Principal Component Analysis) features, model training, and evaluation. These are the basic building blocks of both approaches, then testing them on the same data. I will compare common classical baselines against QML approaches to run on a simulator and will report prediction quality (F1 score). Moreover, practical evaluations like testing multiple encoding techniques under the same experimental conditions and focusing on encoding driven behavior and publishable contribution as it clearly guides which encoding works best for which dataset conditions. The poster will define the main terms for both CML and QML, showing how they are related to explaining the major differences. The project produces clear plots, and guidelines on which encoding choices are most reliable for security style datasets. Concerning education, this can be used as a teaching resource for AI evaluation and emerging QML workflows. In conclusion, this comparative analysis for a small AI and Security task (phishing detection using a tabular dataset of URL/email features) demonstrates strong skills in machine learning, and emerging QML tools that are valuable for research and development.
This is a planned study that compares how much does data encoding techniques affect Quantum Machine Learning (QML) performance compared to Classical Machine Learning (CML)? Unlike classical models, QML cannot directly read a dataset; it must first convert features into quantum states using conversion methods like Classical machine learning reads normal numbers directly, but Quantum machine learning first turns those numbers into quantum form (qubits) that can be present in multiple states at once. This strengthens core AI skills for data preprocessing, PCA (Principal Component Analysis) features, model training, and evaluation. These are the basic building blocks of both approaches, then testing them on the same data. I will compare common classical baselines against QML approaches to run on a simulator and will report prediction quality (F1 score). Moreover, practical evaluations like testing multiple encoding techniques under the same experimental conditions and focusing on encoding driven behavior and publishable contribution as it clearly guides which encoding works best for which dataset conditions. The poster will define the main terms for both CML and QML, showing how they are related to explaining the major differences. The project produces clear plots, and guidelines on which encoding choices are most reliable for security style datasets. Concerning education, this can be used as a teaching resource for AI evaluation and emerging QML workflows. In conclusion, this comparative analysis for a small AI and Security task (phishing detection using a tabular dataset of URL/email features) demonstrates strong skills in machine learning, and emerging QML tools that are valuable for research and development.
Phishing remains a major cybersecurity threat, yet the application of quantum machine learning (QML) to phishing detection is still at an early stage. This study presents a systematic literature review aimed at providing a concise overview of existing QML-based approaches for phishing detection, identifying methodological trends, limitations, and future research directions. A PRISMA-guided review protocol was applied to peer-reviewed journal and conference articles published between 2021 and 2025, retrieved from major scientific databases. Eligible studies were analyzed in terms of QML models, feature encoding strategies, experimental settings, evaluation metrics, and study quality using an adapted Newcastle–Ottawa Scale. The results indicate that current research is limited in volume and largely focuses on hybrid quantum–classical models, particularly quantum support vector machines and variational quantum classifiers. Reported performance is highly dependent on encoding methods, circuit depth, and simulator-based experimentation, with few studies evaluating real quantum hardware. Common challenges include small datasets, lack of external validation, hardware noise, scalability constraints, and the absence of standardized benchmarks. Overall, the review suggests that QML for phishing detection remains exploratory and is not yet competitive with mature classical approaches, but it holds potential as an experimental research direction, provided that future studies address robustness, reproducibility, and practical deployment constraints.
In this paper, we apply Quantum Machine Learning to analyze security datasets. We compare cross-models, Quantum Machine Learning (QML) against Classical Machine Learning (CML), performance with increasing data size, and performance with increasing iteration numbers using commonly used machine learning techniques such as Neural Networks (NN), Support Vector Machines (SVM), and Logistic Regression (LR). Our study focuses on assessing the accuracy of QML and CML approaches on real-world security datasets. The results provide light on the advantages and disadvantages of both QML and CML methodologies, with implications for their use in security data analysis. The experimental findings provide useful information on the applicability of QML and CML for security-related applications. The study contributes to the growing field of quantum machine learning research, particularly in the context of security data analysis, and offers helpful advice for academics and practitioners working in this area.
This study investigates the synergy between Classical Machine Learning (CML) and Quantum Machine Learning (QML) in analyzing security datasets, conducting a comparative analysis using models based on QML and CML to evaluate their performance as data sizes and iteration counts increase.The author, specifically, employs popular machine learning methods, including Support Vector Machines (SVM), Neural Networks (NN), and Logistic Regression (LR), to assess these techniques on real-world security datasets, such as network intrusion detection data and malware classification logs.The primary focus is determining the effectiveness and efficiency of QML and CML approaches in handling large-scale security data.Through rigorous experimentation, the study highlights the benefits and drawbacks of both QML and CML, indicating that while QML offers significant speedups in processing times for large datasets due to quantum parallelism, it faces challenges in terms of hardware accessibility and noise sensitivity, while CML methods, though slower with massive data, benefit from mature algorithms and more robust infrastructure.The outcomes provide critical insights into the practicality of applying QML and CML to security-related applications, demonstrating that QML techniques can outperform CML in specific scenarios, such as real-time threat detection, due to their superior computational efficiency.However, the current limitations of quantum hardware suggest that CML remains more practical for many applications in the short term.This work significantly advances the state of the art in Quantum Machine Learning.It offers vital guidance for practitioners and researchers in security data analysis, underscoring the potential of QML to revolutionize security data processing while acknowledging the ongoing need for advancements in quantum computing technology.
Phishing is a kind of cybercrime where attackers trick unsuspecting network users into revealing sensitive information, resulting in identity theft and financial damage. As this threat continues to grow, artificial intelligence strategies have emerged as a promising solution for its detection in recent years. However, conventional machine learning approaches are starting to show limitations because training on large datasets in standard computing environments may either take too long for accurate results or produce poor accuracy with quick training. The recent advances in the study of the application of quantum computing to machine learning tasks have demonstrated that existing solutions can be strengthened by utilizing the synergy of the two fields, known as Quantum Machine Learning (QML). To explore the potential of this emerging technique, this study implements various QML models integrated with the latest version of Qiskit and presents a detailed execution process along with a comprehensive comparative analysis. The quantitative results obtained from the experiments demonstrate the significant potential of QML models in phishing detection. Furthermore, these state-of-the-art assessments also contribute to future research in improving the composition of QML models applied in the cyber defense field.
The constant rise of malicious URLs continues to pose significant threats and challenges in cybersecurity, with attackers increasingly evading classical detection methods like blacklists and heuristic-based systems. While machine learning (ML) techniques such as SVMs and CNNs have improved detection, their accuracy and scalability remain limited for emerging adversarial approaches. Quantum machine learning (QML) is a transformative strategy that relies on quantum computation and high-dimensional feature spaces to potentially overcome classical computational limitations. However, the accuracy of QML models such as QSVM and QCNN for URL detection in comparison to classical models remains unexplored. This study evaluates ML models (SVMs and CNNs) and QML models (QSVMs and QCNNs) on a dataset, employing data preprocessing techniques such as outliers, feature scaling and feature selection with ANOVA and PCA. Quantum models utilized ZZFeatureMap and ZFeatureMap for data encoding, to transfer original data to qubits. The achieved results showed that CNNs outperformed QCNNs and QSVMs outperformed SVMs in the performance evaluation, demonstrating a competitive potential of quantum computing. QML shows promise for cybersecurity, particularly given the QSVM’s kernel advantages, but current hardware limits the QCNN’s practicality. The significance of this research is to contribute to the growing body of knowledge in cybersecurity by providing a comparative analysis of classical and quantum ML models for classifying malicious URLs.
Quantum Machine Learning (QML) provides a pathway for analysing nonlinear relationships through quantum- enhanced feature spaces. Although quantum models are theoreti- cally well suited for high-complexity data, current hardware and simulator limitations restrict their use to small datasets. For this reason, the original large Web Phishing Detection Dataset was reduced to 200 representative samples through dimensionality reduction, allowing controlled experimentation and enabling evaluation of quantum behaviour with manageable computational cost. QML research commonly follows this approach, testing algorithms on small datasets first, because quantum feature maps can generate exponentially expressive Hilbert-space embeddings even from limited inputs. In this study, multiple Quantum Support Vector Machine (QSVM) models were implemented using fidelity-based quantum kernels and Pauli feature maps. Models with 2 to 10 qubits were evaluated to observe how circuit width affects performance. The two-qubit QSVM achieved the best quantum result with 72 % accuracy, whereas deeper circuits showed declining performance, indicating overfitting and reduced generalization on limited data. In contrast, the classical SVM attained 97 % accuracy with significantly lower computational effort. These findings highlight the challenges and potential of QML for cybersecurity tasks. Future work will explore circuit- expressivity tuning, optimized quantum kernels, hybrid quantum- classical approaches, and scalability to larger phishing datasets as quantum hardware evolves.