Adversarial Feature Selection against Evasion Attacks

Abstract

Pattern recognition and machine learning techniques have been increasingly adopted in adversarial settings such as spam, intrusion and malware detection, although their security against well-crafted attacks that aim to evade detection by manipulating data at test time has not yet been thoroughly assessed. While previous work has been mainly focused on devising adversary-aware classification algorithms to counter evasion attempts, only few authors have considered the impact of using reduced feature sets on classifier security against the same attacks. An interesting, preliminary result is that classifier security to evasion may be even worsened by the application of feature selection. In this paper, we provide a more detailed investigation of this aspect, shedding some light on the security properties of feature selection against evasion attacks. Inspired by previous work on adversary-aware classifiers, we propose a novel adversary-aware feature selection model that can improve classifier security against evasion attacks, by incorporating specific assumptions on the adversary's data manipulation strategy. We focus on an efficient, wrapper-based implementation of our approach, and experimentally validate its soundness on different application examples, including spam and malware detection.

References in corpus (3)

• Poisoning Attacks against Support Vector Machines
• Security Evaluation of Pattern Classifiers under Attack
• Feature Selection via Probabilistic Outputs

Cited by in corpus (8)

• MAD-GAN: Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks
• Black-Box Attacks against RNN based Malware Detection Algorithms
• On Security and Sparsity of Linear Classifiers for Adversarial Settings
• Easy to Fool? Testing the Anti-evasion Capabilities of PDF Malware Scanners
• Automated Poisoning Attacks and Defenses in Malware Detection Systems: An Adversarial Machine Learning Approach
• Smoke Screener or Straight Shooter: Detecting Elite Sybil Attacks in User-Review Social Networks
• Advanced Evasion Attacks and Mitigations on Practical ML-Based Phishing Website Classifiers
• Adequacy of the Gradient-Descent Method for Classifier Evasion Attacks