What question did this study set out to answer?

To develop an efficient visualization-based deep learning system capable of classifying evasive malware and addressing class imbalance without manual feature engineering.

March 15, 2021SHILAP Revista de lepidopterologíaOpen Access

An Efficient DenseNet-Based Deep Learning Model for Malware Detection

View Full Paper

Ask AI

Bookmark

Authors

JHJ. HemalathaAAA College of Engineering and Technology SRS. Abijah RoselineSRM Institute of Science and Technology SGS. GeethaJain University

Discussion

Member takes

Implication

Experimental evaluation demonstrates over 98% classification accuracy across imbalanced malware datasets, highlighting robust defense against evasive and obfuscated attacks.

Key Points

To develop an efficient visualization-based deep learning system capable of classifying evasive malware and addressing class imbalance without manual feature engineering.
Converted binary malware files into two-dimensional images to enable automated spatial feature classification.
Integrated a reweighted class-balanced loss function into the classification layer of a DenseNet model to compensate for underrepresented malware families.
Assessed classification performance and computational overhead across four benchmark datasets: Malimg, BIG 2015, MaleVis, and the unseen Malicia dataset.
Achieved classification accuracies of 98.23% on Malimg, 98.46% on BIG 2015, and 98.21% on MaleVis with reduced false-positive rates and low computational runtime.
Attained 89.48% accuracy on the unseen Malicia dataset, demonstrating high generalizability and resilience against obfuscation attacks.

Cite This Study

Hemalatha et al. (2021) studied this question.

synapsesocial.com/papers/69dd20427a14526fc01335ee https://doi.org/10.3390/e23030344

View Full Paper

Ask AI

Bookmark

Implication

Experimental evaluation demonstrates over 98% classification accuracy across imbalanced malware datasets, highlighting robust defense against evasive and obfuscated attacks.

An Efficient DenseNet-Based Deep Learning Model for Malware Detection

Discussion

Member takes

Implication

Key Points

Cite This Study

An Efficient DenseNet-Based Deep Learning Model for Malware Detection

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider