量子神经网络利用噪声实现鲁棒的入侵检测 · arXiv 研究

作者 PulseAugur 编辑部 · [2 个来源] · 2026-06-23 07:06

本文为随机量子神经网络（SQNN）引入了一个严格的理论框架，以增强网络入侵检测中的对抗鲁棒性。该研究提出了一种“退相干收缩定理”，该定理量化了噪声（特别是退偏信道）如何收缩对抗性扰动。在 NSL-KDD 数据集上的实验表明，与无噪声模型相比，使用这种噪声训练的 SQNN 在抵抗 FGSM 和 PGD 等攻击方面具有显著更强的鲁棒性，并避免了灾难性的鲁棒性崩溃。该研究还推导了噪声正则化的自适应惩罚公式，比较了门控丢弃与退偏噪声，发现它们在减小训练-测试差距方面在统计上没有区别。 AI

影响引入了一种利用量子原理增强人工智能系统对抗鲁棒性的新颖方法，可能带来更安全的人工智能应用。

排序理由该集群包含一篇发表在 arXiv 上的研究论文，详细介绍了量子神经网络的理论和实验发现。

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.CL TIER_1 English(EN) · Gautier-Edouard Edouard Filardo (CREOGN) · 2026-06-24 04:00

Decoherence as Defence and the Magnitude of Noise Regularisation: A Rigorous N -Qubit Theory of Stochastic Quantum Neural Networks for Adversarially Robust Network Intrusion Detection

arXiv:2606.24219v1 Announce Type: new Abstract: Stochastic quantum neural networks (SQNNs) encode neuronal activations as qubits, synaptic topology as entanglement, and neural noise through a Lindblad master equation. A recent conference study applied a ring-entangled SQNN to col…
arXiv cs.CL TIER_1 English(EN) · Gautier-Edouard Edouard Filardo · 2026-06-23 07:06

Decoherence as Defence and the Magnitude of Noise Regularisation: A Rigorous N -Qubit Theory of Stochastic Quantum Neural Networks for Adversarially Robust Network Intrusion Detection

Stochastic quantum neural networks (SQNNs) encode neuronal activations as qubits, synaptic topology as entanglement, and neural noise through a Lindblad master equation. A recent conference study applied a ring-entangled SQNN to collaborative intrusion detection and reached three…

报道来源 [2]

Decoherence as Defence and the Magnitude of Noise Regularisation: A Rigorous N -Qubit Theory of Stochastic Quantum Neural Networks for Adversarially Robust Network Intrusion Detection

Decoherence as Defence and the Magnitude of Noise Regularisation: A Rigorous N -Qubit Theory of Stochastic Quantum Neural Networks for Adversarially Robust Network Intrusion Detection

相关实体

相关话题