标题： 用于油污检测的机器学习量子贝叶斯网络 显示英文标题

标题： Quantum Bayesian Networks for Machine Learning in Oil-Spill Detection

摘要： 量子机器学习（QML）在环境监测、医疗诊断和金融建模等多样化应用中显示出前景。 然而，QML的实际应用面临挑战，例如量子硬件的有限可用性和将量子算法与经典系统集成的复杂性。 本文介绍了一种使用量子贝叶斯网络（QBNs）的新贝叶斯方法，用于分类不平衡数据集，重点是区分卫星衍生数据中的“油污”和“非油污”类别。 通过采用结合概率推理与量子态准备的QBNs，我们有效地解决了将量子增强与经典机器学习架构集成的挑战。 虽然集成提高了关键性能指标，但也揭示了需要改进的领域，突显了针对特定挑战制定定制策略的必要性以优化结果。 我们的研究展示了在检测和分类异常方面的显著进展，有助于更有效和精确的环境监测和管理。 显示英文摘要

摘要： Quantum Machine Learning (QML) has shown promise in diverse applications such as environmental monitoring, healthcare diagnostics, and financial modeling. However, the practical application of QML faces challenges, such as the limited availability of quantum hardware and the complexity of integrating quantum algorithms with classical systems. This paper introduces a novel Bayesian approach using Quantum Bayesian Networks (QBNs) to classify imbalanced datasets, focusing on differentiating ``oil-spill'' from ``non-spill'' classes in satellite-derived data. By employing QBNs, which combine probabilistic reasoning with quantum state preparation, we effectively address the challenge of integrating quantum enhancements with classical machine learning architectures. While the integration improves key performance metrics, it also uncovers areas for refinement, highlighting the need for customized strategies to address specific challenges and optimize outcomes. Our study demonstrates significant advances in detecting and classifying anomalies, contributing to more effective and precise environmental monitoring and management.