Black-Box Attacks against RNN based Malware Detection Algorithms

Abstract

Recent researches have shown that machine learning based malware detection algorithms are very vulnerable under the attacks of adversarial examples. These works mainly focused on the detection algorithms which use features with fixed dimension, while some researchers have begun to use recurrent neural networks (RNN) to detect malware based on sequential API features. This paper proposes a novel algorithm to generate sequential adversarial examples, which are used to attack a RNN based malware detection system. It is usually hard for malicious attackers to know the exact structures and weights of the victim RNN. A substitute RNN is trained to approximate the victim RNN. Then we propose a generative RNN to output sequential adversarial examples from the original sequential malware inputs. Experimental results showed that RNN based malware detection algorithms fail to detect most of the generated malicious adversarial examples, which means the proposed model is able to effectively bypass the detection algorithms.

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Categorical Reparameterization with Gumbel-Softmax
• Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples
• Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN
• Adversarial Perturbations Against Deep Neural Networks for Malware Classification
• Adversarial Feature Selection against Evasion Attacks
• Simple Black-Box Adversarial Perturbations for Deep Networks
• Practical Black-Box Attacks against Machine Learning
• A General Retraining Framework for Scalable Adversarial Classification

Cited by in corpus (21)

• ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Models
• Deep Learning for Anomaly Detection: A Survey
• IDSGAN: Generative Adversarial Networks for Attack Generation against Intrusion Detection
• Black-box Adversarial Attacks with Limited Queries and Information
• Adversarial Attacks on Deep Learning Models in Natural Language Processing: A Survey
• Adversarial Attacks against Windows PE Malware Detection: A Survey of the State-of-the-Art
• Deceiving End-to-End Deep Learning Malware Detectors using Adversarial Examples
• POBA-GA: Perturbation Optimized Black-Box Adversarial Attacks via Genetic Algorithm
• Defense Methods Against Adversarial Examples for Recurrent Neural Networks
• Generation & Evaluation of Adversarial Examples for Malware Obfuscation
• Binary Black-box Evasion Attacks Against Deep Learning-based Static Malware Detectors with Adversarial Byte-Level Language Model
• On the Design of Black-box Adversarial Examples by Leveraging Gradient-free Optimization and Operator Splitting Method
• Generic Black-Box End-to-End Attack Against State of the Art API Call Based Malware Classifiers
• Towards Security Threats of Deep Learning Systems: A Survey
• ATMPA: Attacking Machine Learning-based Malware Visualization Detection Methods via Adversarial Examples
• Semantic-preserving Reinforcement Learning Attack Against Graph Neural Networks for Malware Detection
• HeNet: A Deep Learning Approach on Intel${}^{\circledR }$ Processor Trace for Effective Exploit Detection
• Spatially Focused Attack against Spatiotemporal Graph Neural Networks
• On the Veracity of Cyber Intrusion Alerts Synthesized by Generative Adversarial Networks
• Generative Models for Security: Attacks, Defenses, and Opportunities
• LSDAT: Low-Rank and Sparse Decomposition for Decision-based Adversarial Attack