Skip to main content
. 2022 Mar 19;12:4770. doi: 10.1038/s41598-022-08504-6

Table 1.

Comparison of the classification between the different types of neural networks.

Type Our results Related works Related works results
ANN

Contribution 1:

The binary detection (malware and goodware) gave us an accuracy of 100%

 7 The accuracy for the classification of malware is 97.8%

Contribution 2:

The accuracy for classification (MLP) of nine families of ransomware is 91%

 9

The accuracy for the classification of ransomware:

 Using ANN is ≈ 70%

 Using BN is ≈ 49%
CNN The accuracy for classification of nine families of ransomware is 94% 11 They obtained in the results an accuracy of 98%
RNN The accuracy for classification of nine families of ransomware is 79% 13 No accuracy value
16

 ARI-LSTM (L = 5) = 0.93 (93%)

 ARI-LSTM (L = 8) = 0.91 (91%)

the values in bold are the results of accuracy obtained.