Conformal Prediction for Offensive Security

📅 2026-09-04
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
研究探讨了在进攻性安全领域,特别是隐私保护机器学习和网络流量分析中,如何应用共形预测方法来执行攻击。
📝 Abstract
Despite its introduction more than a quarter century ago, Conformal Prediction (CP) has seen surprisingly few applications to the cyber security world thus far. In particular, we observe that, while CP has been employed as a defensive measure in many recent works, its use for carrying out attacks (i.e., for offensive security) is hard to trace in the literature. We explore this gap, by presenting initial findings in two key areas of offensive security: Privacy-Preserving Machine Learning, and network traffic analysis.
Problem

Research questions and friction points this paper is trying to address.

Conformal Prediction
offensive security
Privacy-Preserving Machine Learning
network traffic analysis
Innovation

Methods, ideas, or system contributions that make the work stand out.

Conformal Prediction
Offensive Security
Privacy-Preserving Machine Learning
Network Traffic Analysis