Research Article

Malicious Code Variant Identification Based on Multiscale Feature Fusion CNNs

Table 3

Comparative summary of MFFC algorithm with previous malware classification algorithms.

MethodYearTechniqueAccuracy (%)Precision (%)Recall (%)F1 score (%)

Nataraj et al. [6]2011ML97.18
SPAM-GIST [10]2016ML97.40
DL + SVM [17]2017DL + ML84.92
Vgg-verydeep-19 [16]2017DL97.32
GIST + SVM [18]2018ML92.2092.5091.40
GIST + KNN [18]2018ML91.9092.1091.70
GLCM + SVM [18]2018ML93.2093.4093.00
GLCM + KNN [18]2018ML92.5092.7092.30
DRBA + CNNs [18]2018DL94.5096.6088.40
LGMP + KNN [11]2019ML98.4098.20
NSGA-II + CNNs [19]2019DL97.6097.6088.40
Venkatraman [22]2019DL96.3091.8091.5091.60
Gibert et al. [28]2019DL98.5098.0098.0098.00
IMCFN [29]2020DL98.8298.8598.8198.75
DEAM-DenseNet [30]2021DL98.5096.9096.6096.70
MFFC2021DL98.7298.8698.7298.73

ML: machine learning; DL: deep learning.