· 2013
Evasion attacks against machine learning at test time
- 1,636citations
- 2013year
Short summary
A gradient-based method systematically assesses machine learning classifier security against test-time evasion attacks, demonstrating that malware detection systems can be easily evaded.
AI-generated from the title and abstract; the full text is not read.
Key points
- A gradient-based method is proposed to systematically assess machine learning classifier security against evasion attacks at test time.
- The method simulates attack scenarios with increasing attacker knowledge and manipulation ability to reveal classifier vulnerabilities.
- The approach was evaluated on PDF malware detection, demonstrating that such systems can be easily evaded.
- The analysis also suggests potential countermeasures against these evasion attacks.
AI-generated from the title and abstract; the full text is not read.
Abstract
Abstract. In security-sensitive applications, the success of machine learn-ing depends on a thorough vetting of their resistance to adversarial data. In one pertinent, well-motivated attack scenario, an adversary may at-tempt to evade a deployed system at test time by carefully manipulating attack samples. In this work, we present a simple but effective gradient-based approach that can be exploited to systematically assess the security of several, widely-used classification algorithms against evasion attacks. Following a recently proposed framework for security evaluation, we sim-ulate attack scenarios that exhibit different risk levels for the classifier by increasing the attacker’s knowledge of the system and her ability to manipulate attack samples. This gives the classifier designer a better pic-ture of the classifier performance under evasion attacks, and allows him to perform a more informed model selection (or parameter setting). We evaluate our approach on the relevant security task of malware detection in PDF files, and show that such systems can be easily evaded. We also sketch some countermeasures suggested by our analysis.
The authors' abstract, as published at the source., 2013 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Signal Processing
Signal ProcessingComputer Science