Table 11 Comparison of AndroMalPack with related work.

From: AndroMalPack: enhancing the ML-based malware classification by detection and removal of repacked apps for Android systems

Technique

ML-Model

Features

Dataset

Accuracy

Recall

Precision

74

Ensemble

Permissions and source code

M0Droid

95.6

95.7

95.8

75

Random Forrest

Permissions, APIs and system calls

Custom

88.26

88.40

88.16

76

Random Forrest

Permissions, APIs,intents and hardware

Drebin

97.24

96.88

97.58

77

Random Forrest

API calls

Drebin

96

95

97

AMD

98

98

99

78

CatBoost

Permissions and op-code sequence

Drebin and Custom

97.40

96.77

98.0

79

Random Forrest

Permissions based features

Androzoo

81.53

81.53

82.59

80

Profile hidden Markov model

API calls

Drebin

94.5

N/A

93.0

81

SVM

Permission, intents and byte-code

AMD

95.09

95.09

95.11

AndroMalPack

Random Forrest

Permission, API calls and Intents

Drebin

98.29

98.70

97.7

AMD

98.21

99.40

98.1

Androzoo

97.14

99.8

97.2