Next Article in Journal
Argumentation-Based Equation-Solver Selector for Energy-Efficient WSNs
Previous Article in Journal
Quantifying System-Level Risk at Highway–Rail Grade Crossings: Integrating Spatial Autocorrelation and Explainable Machine Learning
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Algorithms
Volume 19
Issue 6
10.3390/a19060456
algorithms-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

ProtoMal: Prototype-Guided Dual-Branch Continual Learning for Robust Android Malware Detection

by
Xuan Zhang
1,
Aihua Zhang
1,*,
Maode Ma
2,
Yuanjie Bo
1,
Yiying Zhang
1 and
Yanan Zhang
1
1
Department of Artificial Intelligence, Tianjin University of Science and Technology, Tianjin 300457, China
2
Faculty of Computer Science and Artificial Intelligence, Shenzhen University of Advanced Technology, Shenzhen 518055, China
*
Author to whom correspondence should be addressed.
Algorithms 2026, 19(6), 456; https://doi.org/10.3390/a19060456
Submission received: 13 April 2026 / Revised: 19 May 2026 / Accepted: 26 May 2026 / Published: 4 June 2026
(This article belongs to the Section Evolutionary Algorithms and Machine Learning)
Versions Notes

Abstract

Traditional Android malware detection systems struggle to adapt to evolving threats without sacrificing performance on legacy families. To address this, we present ProtoMal, a dual-branch continual learning framework that achieves a fine-grained balance between stability and plasticity. The framework utilizes a frozen old branch for knowledge preservation and a trainable new branch for novel threat acquisition. A key contribution is our robust median-based prototype learning mechanism, which leverages centroids and outlier filtering to handle the high intra-class variability and label noise inherent in malware datasets. Experimental results across three large-scale benchmarks AMD, VirusShare, and VirusShareYears demonstrate that ProtoMal significantly curtails performance degradation and achieves highly competitive average accuracy. Most notably, the proposed framework demonstrates highly competitive model stability and yields robust anti-forgetting capabilities alongside current state-of-the-art incremental learning paradigms, maintaining particular resilience under severe concept drift.
Keywords: malware; class-incremental learning; deep learning malware; class-incremental learning; deep learning

Share and Cite

MDPI and ACS Style

Zhang, X.; Zhang, A.; Ma, M.; Bo, Y.; Zhang, Y.; Zhang, Y. ProtoMal: Prototype-Guided Dual-Branch Continual Learning for Robust Android Malware Detection. Algorithms 2026, 19, 456. https://doi.org/10.3390/a19060456

AMA Style

Zhang X, Zhang A, Ma M, Bo Y, Zhang Y, Zhang Y. ProtoMal: Prototype-Guided Dual-Branch Continual Learning for Robust Android Malware Detection. Algorithms. 2026; 19(6):456. https://doi.org/10.3390/a19060456

Chicago/Turabian Style

Zhang, Xuan, Aihua Zhang, Maode Ma, Yuanjie Bo, Yiying Zhang, and Yanan Zhang. 2026. "ProtoMal: Prototype-Guided Dual-Branch Continual Learning for Robust Android Malware Detection" Algorithms 19, no. 6: 456. https://doi.org/10.3390/a19060456

APA Style

Zhang, X., Zhang, A., Ma, M., Bo, Y., Zhang, Y., & Zhang, Y. (2026). ProtoMal: Prototype-Guided Dual-Branch Continual Learning for Robust Android Malware Detection. Algorithms, 19(6), 456. https://doi.org/10.3390/a19060456

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop