Efficient Query-Based Attack against ML-Based Android Malware Detection under Zero Knowledge Setting

Published in

Conference on Computer and Communications Security(2023)

External Links:

Generate Graph DownloadPDF

TL;DR

AdvDroidZero is introduced, an efficient query-based attack framework against ML-based AMD methods that operates under the zero knowledge setting that is effective against various mainstream ML- based AMD methods, in particular, state-of-the-art such methods and real-world antivirus solutions.

Abstract

The widespread adoption of the Android operating system has made malicious Android applications an appealing target for attackers. Machine learning-based (ML-based) Android malware detection (AMD) methods are crucial in addressing this problem; however, their vulnerability to adversarial examples raises concerns. Current attacks against ML-based AMD methods demonstrate remarkable performance but rely on strong assumptions that may not be realistic in real-world scenarios, e.g., the knowledge requirements about feature space, model parameters, and training dataset. To address this limitation, we introduce AdvDroidZero, an efficient query-based attack framework against ML-based AMD methods that operates under the zero knowledge setting. Our extensive evaluation shows that AdvDroidZero is effective against various mainstream ML-based AMD methods, in particular, state-of-the-art such methods and real-world antivirus solutions.

Authors

Ping He

1 papers

Yifan Xia

1 papers

Xuhong Zhang

1 papers

References74 items

Black-box Adversarial Example Attack towards FCG Based Android Malware Detection under Incomplete Feature Information

“Real Attackers Don't Compute Gradients”: Bridging the Gap Between Adversarial ML Research and Practice

Towards Certifying the Asymmetric Robustness for Neural Networks: Quantification and Applications

Statista

MAB-Malware: A Reinforcement Learning Framework for Blackbox Generation of Adversarial Malware

Efficient Query-Based Attack against ML-Based Android Malware Detection under Zero Knowledge Setting

Published in

Conference on Computer and Communications Security(2023)

External Links:

Generate Graph DownloadPDF

TL;DR

Abstract

Authors

Ping He

1 papers

Yifan Xia

1 papers

Xuhong Zhang

1 papers

References74 items

Black-box Adversarial Example Attack towards FCG Based Android Malware Detection under Incomplete Feature Information

“Real Attackers Don't Compute Gradients”: Bridging the Gap Between Adversarial ML Research and Practice

Towards Certifying the Asymmetric Robustness for Neural Networks: Quantification and Applications

Statista

MAB-Malware: A Reinforcement Learning Framework for Blackbox Generation of Adversarial Malware

Shouling Ji

1 papers

A Deep Dive Inside DREBIN: An Explorative Analysis beyond Android Malware Detection Scores

Transfer Attacks Revisited: A Large-Scale Empirical Study in Real Computer Vision Settings

RamBoAttack: A Robust Query Efficient Deep Neural Network Decision Exploit

ShadowDroid: Practical Black-box Attack against ML-based Android Malware Detection

Structural Attack against Graph Based Android Malware Detection

Cert-RNN: Towards Certifying the Robustness of Recurrent Neural Networks

EvadeDroid: A Practical Evasion Attack on Machine Learning for Black-box Android Malware Detection

A Large-scale Temporal Measurement of Android Malicious Apps: Persistence, Migration, and Lessons Learned

Robust Android Malware Detection against Adversarial Example Attacks

Deep Learning for Android Malware Defenses: A Systematic Literature Review

PyCG: Practical Call Graph Generation in Python

Understanding Worldwide Private Information Collection on Android

Enhancing State-of-the-art Classifiers with API Semantics to Detect Evolved Android Malware

Transcending TRANSCEND: Revisiting Malware Classification in the Presence of Concept Drift

SoK: Certified Robustness for Deep Neural Networks

Blacklight: Scalable Defense for Neural Networks against Query-Based Black-Box Attacks

Adversarial Deep Ensemble: Evasion Attacks and Defenses for Malware Detection

Why an Android App is Classified as Malware? Towards Malware Classification Interpretation

Improving Certified Robustness via Statistical Learning with Logical Reasoning

Mind Your Weight(s): A Large-scale Study on Insufficient Machine Learning Model Protection in Mobile Apps

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Intriguing Properties of Adversarial ML Attacks in the Problem Space

Hybrid Batch Attacks: Finding Black-box Adversarial Examples with Limited Queries

HopSkipJumpAttack: A Query-Efficient Decision-Based Attack

On Evaluating Adversarial Robustness

TextBugger: Generating Adversarial Text Against Real-world Applications

Interpretable Deep Learning under Fire

LEMNA: Explaining Deep Learning based Security Applications

Android HIV: A Study of Repackaging Malware for Evading Machine-Learning Detection

TESSERACT: Eliminating Experimental Bias in Malware Classification across Space and Time

The Spyware Used in Intimate Partner Violence

When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks

Eight Years of Rider Measurement in the Android Malware Ecosystem

Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models

Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning

Malware Detection in Adversarial Settings: Exploiting Feature Evolutions and Confusions in Android Apps

Adversarial Examples for Malware Detection

Yes, Machine Learning Can Be More Secure! A Case Study on Android Malware Detection

MaMaDroid: Detecting Android Malware by Building Markov Chains of Behavioral Models

A Boundary Tilting Persepective on the Phenomenon of Adversarial Examples

Towards Evaluating the Robustness of Neural Networks

AndroZoo: Collecting Millions of Android Apps for the Research Community

Practical Black-Box Attacks against Machine Learning

Dead code elimination technique in eclipse compiler for Java

LOOP: Logic-Oriented Opaque Predicate Detection in Obfuscated Binary Code

Android Security: A Survey of Issues, Malware Penetration, and Defenses

FlowDroid: precise context, flow, field, object-sensitive and lifecycle-aware taint analysis for Android apps

Intriguing properties of neural networks

PScout: analyzing the Android permission specification

Soot: a Java bytecode optimization framework

Limits of Static Analysis for Malware Detection

Opaque Predicates Detection by Abstract Interpretation

An introduction to decision tree modeling

An empirical study of static call graph extractors

Learning decision tree classifiers

2023. Android Permission

2023. Total Amount of Android Malware

AutoDA: Automated Decision-based Iterative Adversarial Attacks

Dos and Don'ts of Machine Learning in Computer Security

Exposing Weaknesses of Malware Detectors with Explainability-Guided Evasion Attacks

CADE: Detecting and Explaining Concept Drift Samples for Security Applications

Transcend: Detecting Concept Drift in Malware Classification Models

DREBIN: Effective and Explainable Detection of Android Malware in Your Pocket

This paper is included in the Proceedings of the 30th USENIX Security Symposium.

Use-Feature reference

Field of Study

Computer Science

Journal Information

Name

Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security

Venue Information

Name

Conference on Computer and Communications Security

Type

conference

URL

https://dl.acm.org/conference/ccs

Alternate Names

Int Workshop Cogn Cell Syst
CCS
Comput Commun Secur
CcS
International Symposium on Community-centric Systems
International Workshop on Cognitive Cellular Systems
Conf Comput Commun Secur
Comb Comput Sci
Int Symp Community-centric Syst
Combinatorics and Computer Science
Circuits, Signals, and Systems
Computer and Communications Security
Circuit Signal Syst

TL;DR

Abstract

Authors

References74 items

Black-box Adversarial Example Attack towards FCG Based Android Malware Detection under Incomplete Feature Information

“Real Attackers Don't Compute Gradients”: Bridging the Gap Between Adversarial ML Research and Practice

Towards Certifying the Asymmetric Robustness for Neural Networks: Quantification and Applications

Statista

MAB-Malware: A Reinforcement Learning Framework for Blackbox Generation of Adversarial Malware

TL;DR

Abstract

Authors

References74 items

Black-box Adversarial Example Attack towards FCG Based Android Malware Detection under Incomplete Feature Information

“Real Attackers Don't Compute Gradients”: Bridging the Gap Between Adversarial ML Research and Practice

Towards Certifying the Asymmetric Robustness for Neural Networks: Quantification and Applications

Statista

MAB-Malware: A Reinforcement Learning Framework for Blackbox Generation of Adversarial Malware

A Deep Dive Inside DREBIN: An Explorative Analysis beyond Android Malware Detection Scores

Transfer Attacks Revisited: A Large-Scale Empirical Study in Real Computer Vision Settings

RamBoAttack: A Robust Query Efficient Deep Neural Network Decision Exploit

ShadowDroid: Practical Black-box Attack against ML-based Android Malware Detection

Structural Attack against Graph Based Android Malware Detection

Cert-RNN: Towards Certifying the Robustness of Recurrent Neural Networks

EvadeDroid: A Practical Evasion Attack on Machine Learning for Black-box Android Malware Detection

A Large-scale Temporal Measurement of Android Malicious Apps: Persistence, Migration, and Lessons Learned

Robust Android Malware Detection against Adversarial Example Attacks

Deep Learning for Android Malware Defenses: A Systematic Literature Review

PyCG: Practical Call Graph Generation in Python

Understanding Worldwide Private Information Collection on Android

Enhancing State-of-the-art Classifiers with API Semantics to Detect Evolved Android Malware

Transcending TRANSCEND: Revisiting Malware Classification in the Presence of Concept Drift

SoK: Certified Robustness for Deep Neural Networks

Blacklight: Scalable Defense for Neural Networks against Query-Based Black-Box Attacks

Adversarial Deep Ensemble: Evasion Attacks and Defenses for Malware Detection

Why an Android App is Classified as Malware? Towards Malware Classification Interpretation

Improving Certified Robustness via Statistical Learning with Logical Reasoning

Mind Your Weight(s): A Large-scale Study on Insufficient Machine Learning Model Protection in Mobile Apps

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Intriguing Properties of Adversarial ML Attacks in the Problem Space

Hybrid Batch Attacks: Finding Black-box Adversarial Examples with Limited Queries

HopSkipJumpAttack: A Query-Efficient Decision-Based Attack

On Evaluating Adversarial Robustness

TextBugger: Generating Adversarial Text Against Real-world Applications

Interpretable Deep Learning under Fire

LEMNA: Explaining Deep Learning based Security Applications

Android HIV: A Study of Repackaging Malware for Evading Machine-Learning Detection

TESSERACT: Eliminating Experimental Bias in Malware Classification across Space and Time

The Spyware Used in Intimate Partner Violence

When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks

Eight Years of Rider Measurement in the Android Malware Ecosystem

Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models

Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning

Malware Detection in Adversarial Settings: Exploiting Feature Evolutions and Confusions in Android Apps

Adversarial Examples for Malware Detection

Yes, Machine Learning Can Be More Secure! A Case Study on Android Malware Detection

MaMaDroid: Detecting Android Malware by Building Markov Chains of Behavioral Models

A Boundary Tilting Persepective on the Phenomenon of Adversarial Examples

Towards Evaluating the Robustness of Neural Networks

AndroZoo: Collecting Millions of Android Apps for the Research Community

Practical Black-Box Attacks against Machine Learning

Dead code elimination technique in eclipse compiler for Java

LOOP: Logic-Oriented Opaque Predicate Detection in Obfuscated Binary Code

Android Security: A Survey of Issues, Malware Penetration, and Defenses

FlowDroid: precise context, flow, field, object-sensitive and lifecycle-aware taint analysis for Android apps

Intriguing properties of neural networks

PScout: analyzing the Android permission specification

Soot: a Java bytecode optimization framework

Limits of Static Analysis for Malware Detection

Opaque Predicates Detection by Abstract Interpretation

An introduction to decision tree modeling

An empirical study of static call graph extractors

Learning decision tree classifiers

AV-ATLAS

2023. Android Permission

VirusTotal

2023. Total Amount of Android Malware

AutoDA: Automated Decision-based Iterative Adversarial Attacks

Dos and Don'ts of Machine Learning in Computer Security

Exposing Weaknesses of Malware Detectors with Explainability-Guided Evasion Attacks