Adversarial Attacks on Transformers-Based Malware Detectors

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jakhotiya, Yash, Patil, Heramb, Rawlani, Jugal, Mane, Sunil B.
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916374964600832
author Jakhotiya, Yash
Patil, Heramb
Rawlani, Jugal
Mane, Sunil B.
author_facet Jakhotiya, Yash
Patil, Heramb
Rawlani, Jugal
Mane, Sunil B.
contents Signature-based malware detectors have proven to be insufficient as even a small change in malignant executable code can bypass these signature-based detectors. Many machine learning-based models have been proposed to efficiently detect a wide variety of malware. Many of these models are found to be susceptible to adversarial attacks - attacks that work by generating intentionally designed inputs that can force these models to misclassify. Our work aims to explore vulnerabilities in the current state of the art malware detectors to adversarial attacks. We train a Transformers-based malware detector, carry out adversarial attacks resulting in a misclassification rate of 23.9% and propose defenses that reduce this misclassification rate to half. An implementation of our work can be found at https://github.com/yashjakhotiya/Adversarial-Attacks-On-Transformers.
format Preprint
id arxiv_https___arxiv_org_abs_2210_00008
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Adversarial Attacks on Transformers-Based Malware Detectors
Jakhotiya, Yash
Patil, Heramb
Rawlani, Jugal
Mane, Sunil B.
Cryptography and Security
Artificial Intelligence
Machine Learning
Signature-based malware detectors have proven to be insufficient as even a small change in malignant executable code can bypass these signature-based detectors. Many machine learning-based models have been proposed to efficiently detect a wide variety of malware. Many of these models are found to be susceptible to adversarial attacks - attacks that work by generating intentionally designed inputs that can force these models to misclassify. Our work aims to explore vulnerabilities in the current state of the art malware detectors to adversarial attacks. We train a Transformers-based malware detector, carry out adversarial attacks resulting in a misclassification rate of 23.9% and propose defenses that reduce this misclassification rate to half. An implementation of our work can be found at https://github.com/yashjakhotiya/Adversarial-Attacks-On-Transformers.
title Adversarial Attacks on Transformers-Based Malware Detectors
topic Cryptography and Security
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2210.00008