Syscall-based HIDS generalization to unseen CVEs shows mixed results

Researchers explored the generalization capabilities of system-call based Host Intrusion Detection Systems (HIDS). They investigated whether a HIDS trained on normal behavior for specific Common Vulnerabilities and Exposures (CVEs) sharing a Common Weakness Enumeration (CWE) class could detect unseen CVEs within the same class. The study found that while CWE-level generalization is achievable for some weakness families, it is not universal and is heavily influenced by the breadth of the normal behavior profile rather than solely the CWE label. AI

IMPACT This research explores the limitations and potential of generalization in security systems, which could inform the development of more robust AI-driven intrusion detection.

RANK_REASON Academic paper detailing a research study on HIDS generalization. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Hugging Face Daily Papers →

• Host intrusion detection systems
• Isolation Forest
• SGD One-Class SVM

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →