|
| Reference | Methodology | Dataset | Result |
|
| [50] | STL | NSL-KDD | STL and SM have precision values of 85.44% and 96.56%, respectively. On the other hand, the STL outperformed SM in terms of recall. STL and SM had recall values of 95.95% and 63.73%, respectively |
| [51] | Deep convolutional neural network (DCNN) | KDD CUP 99 | Accuracy of 99%
False alarm of 0.08% |
| [52] | The DBN (stacking multiple RBMs) | KDD CUP 99 | Accuracy rate of 95%, a low false-negative rate of 2.48%, and a high true-positive rate of 97.5% |
| [53] | Deep convolutional neural network (DCNN) | KDD CUP 99 | Accuracy of 99.89% |
| [54] | Deep convolutional neural network (DCNN) | NSL-KDD | Accuracy of 98.90%. |
| [55] | LSTM and DBN | NSL-KDD | 74.188%, 71.30%, 71.91, and 73.18% for RF, SVM, DBN, and LSTM, respectively, for KDDTest+ and 51.02%, 45.54%, 46.73, and 49.37% for KDDTest-21 |
| [56] | Deep neural network, which consists of sparse autoencoder and logistic regression. By stacking the autoencoders, a deep network is built, and a logistic regression network is used to classify the features learned | NSL-KDD | Precision was 84.6%, and its recall score was 92.8%. The specificity and negative predictive values were 80.7% and 90.7%, respectively. The model’s overall accuracy was 87.2% |
| [57] | Deep convolutional neural network (DCNN) | CSE-CIC-IDS 2018 | In CNN, the detection rate of benign, DoS, brute force, SQL injection attack, and infiltration attack was 1, 0.97, 0.86, 0.57, and 0.33, respectively |
| [58] | Deep convolutional neural network (DCNN) | CICIDS2017 dataset | 99.95% overall accuracy, a precision of 94.31%, a recall or detection rate of 95.62%, and an F1 score of 94.1% |
| [59] | Deep convolutional neural network (DCNN) | KDD CUP 99 | The model’s accuracy is greater than 99% |
| [60] | SVM and deep convolutional neural network (DCNN) | NSL-KDD | Accuracy of 97% |
| [61] | Deep convolutional neural network (DCNN) | KDD 99 | Accuracy of 97.7% |
| [62] | Deep CNN | NSL-KDD | Accuracy of 95.45% |
| [63] | LSTM | UNSW NB15 dataset | Accuracy of 98% |
| [64] | Deep CNN | CICIDS2017 dataset | The detection rate was 96.55% |
|
