|
Jiang et al. (2019)
|
2019 |
Opcode, Sensitive API, STR, actions |
Drebin (Android) |
KNN, DT, GB |
99.5% (AUC) |
|
Sung et al. (2020)
|
2020 |
Opcode sequence as text |
BIG 2015 (Windows) |
Bi-LSTM, fastText |
96.7% |
|
Zou et al. (2020)
|
2020 |
ByteCode |
Drebin (Android) |
CNN |
97% |
|
Jeon & Moon (2020)
|
2020 |
Opcode sequence |
Windows hybrid own dataset (Windows) |
C-Autoenconder+RNN |
96% |
|
Singh et al. (2020)
|
2020 |
Opcode, Permissions, Intents, LSI |
CICInvesAndMal2019 (Android) |
RF |
93.92% |
|
Niu et al. (2020)
|
2020 |
Opcode, CFG |
Hybrid own dataset (Android) |
LSTM |
97% |
|
Bai et al. (2020)
|
2020 |
Opcode, N-gram (N = 5), permissions |
Drebin (Android) |
CatBoost |
96.21% |
|
Zhang et al. (2019)
|
2020 |
Opcode- Bi-gram, API Calls |
Hybrid own dataset (Windows) |
CNN, BPNN |
95% |
|
Parildi, Hatzinakos & Lawryshyn (2021)
|
2021 |
OpCode, Word2Vec |
Hybrid own dataset (Windows) |
1-Katman: CNN, LSTM |
95% |
| 2-Katman: LR, SVM, KNN, RF, GB |
|
Bhat & Dutta (2021)
|
2021 |
Opcode-Ngram |
PRAGuard+Google play store (Android) |
Algoritmic solution |
96.22% |
|
Sihag, Vardhan & Singh (2021)
|
2021 |
Opcode sequence |
Drebin+Genome+Contagio+AndroAutopssy+AndroDracker (Android) |
KNN, J48, RF |
98.18% |
|
Darem et al. (2021)
|
2021 |
Opcode, N-gram(1, 2, 3), Image conversion |
Microsoft malware dataset (Windows) |
XGBoost and CNN (Ensemble) |
99.12% |
|
Zhang et al. (2021)
|
2021 |
Opcode sequence, System call, Text |
Drebin (Android) |
CNN-BiLSTM, NLP |
97.6% |
|
Khan et al. (2022)
|
2022 |
Opcode, Skip-gram |
Own hybrid dataset (Android) |
Op2Vec, DNN, |
97.47% |
|
Wang & Qian (2022)
|
2022 |
Opcode, Word2Vec |
SOREL-20M (Windows) |
TextCNN |
98.66% |
|
Jeon et al. (2022)
|
2022 |
Opcode sequence, API calls, RGB Images (Hybrid) |
HyMalD (Android) |
HyMaID (Bi-LSTM, SPP-NET) |
92.5% |
|
Tang et al. (2023)
|
2023 |
Bytecode, Hex, N-gram |
BIG 2015 and Malimg (Windows) |
LightGBM |
99.70% |
|
Xie, Qin & Di (2023)
|
2023 |
Dalvik opcode, Permissions, API calls |
CIC-AndMal2017 and CICMalDroid2020 (Android) |
Stacking Model |
96.77% |
|
Wu et al. (2023)
|
2023 |
Opcode, Function-calls, graph2vec |
Own hybrid dataset (IOT) |
SVM |
98.88% |
|
Lee et al. (2023)
|
2023 |
Opcode |
Own hybrid dataset (Android) |
Multilayer perceptron (MLP) |
98.0% |
