Revolutionizing Obfuscated Malware Detection Through Memory-Aware Hybrid Attention Networks (MAHAN)

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62690094%3A18450%2F25%3A50023281" target="_blank" >RIV/62690094:18450/25:50023281 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1109/ICOCO67189.2025.11334091" target="_blank" >http://dx.doi.org/10.1109/ICOCO67189.2025.11334091</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICOCO67189.2025.11334091" target="_blank" >10.1109/ICOCO67189.2025.11334091</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Revolutionizing Obfuscated Malware Detection Through Memory-Aware Hybrid Attention Networks (MAHAN)

Popis výsledku v původním jazyce

Recently, obfuscated malware can be considered as one of the big problems for cybersecurity since many research studies were conducted to overcome it. Thus, this research presents an innovative deep learning architecture called Memory-Aware Hybrid Attention Network (MAHAN). It helps to assist researchers in finding obfuscated malware in different ways. The CIC-MalMem-2022 dataset is used in a full benchmarking analysis to compare deep learning models such as CNN, LSTM, and CNN-LSTM with traditional machine learning models such as LightGBM, Random Forest, XGBoost, K-Nearest Neighbours, Support Vector Machine, and Naive Bayes. MAHAN uses dilated convolutions, bidirectional LSTM, and multi-head attention to find complex patterns in volatile memory data which change over time and space. The experimental results show that MAHAN is competitive and either beats or matches the best models in detecting the obfuscated malware. Our findings demonstrate that memory-aware, and attention-based architectures may facilitate the detection of obfuscated malware, particularly when the malwares are not easily perceptible. © 2025 IEEE.

Název v anglickém jazyce

Revolutionizing Obfuscated Malware Detection Through Memory-Aware Hybrid Attention Networks (MAHAN)

Popis výsledku anglicky

Recently, obfuscated malware can be considered as one of the big problems for cybersecurity since many research studies were conducted to overcome it. Thus, this research presents an innovative deep learning architecture called Memory-Aware Hybrid Attention Network (MAHAN). It helps to assist researchers in finding obfuscated malware in different ways. The CIC-MalMem-2022 dataset is used in a full benchmarking analysis to compare deep learning models such as CNN, LSTM, and CNN-LSTM with traditional machine learning models such as LightGBM, Random Forest, XGBoost, K-Nearest Neighbours, Support Vector Machine, and Naive Bayes. MAHAN uses dilated convolutions, bidirectional LSTM, and multi-head attention to find complex patterns in volatile memory data which change over time and space. The experimental results show that MAHAN is competitive and either beats or matches the best models in detecting the obfuscated malware. Our findings demonstrate that memory-aware, and attention-based architectures may facilitate the detection of obfuscated malware, particularly when the malwares are not easily perceptible. © 2025 IEEE.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2025 IEEE International Conference on Computing, ICOCO 2025

ISBN

979-8-3315-7539-7

ISSN

—

e-ISSN

—

Počet stran výsledku

6

Strana od-do

453-458

Název nakladatele

Institute of Electrical and Electronics Engineers Inc.

Místo vydání

New York

Místo konání akce

Kuching

Datum konání akce

6. 10. 2025

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—