Research Article
Neural Network-Based Voting System with High Capacity and Low Computation for Intrusion Detection in SIEM/IDS Systems
Table 7
Comparison with other approaches.
| | Approach | Model capacity | Computational cost | Accuracy (%) | Precision (%) | F-score (%) | Recall (%) |
| | NN with best cost and training function [9] | Low | Low | 81.8 | 92.5 | — | 67.9 | | Sparse autoencoder with SMR [10] | Low | High | 78.06 | 96.56 | 76.8 | 63.73 | | Sparse autoencoder with logistic regression [11] | High | High | 87.2 | 84.6 | — | 92.8 | | RNN based IDS [12] | Low | High | 81.29 | — | — | — | | CBR-CNN based IDS [13] | High | High | 89.41 | 92.63 | — | — | | This research | High | Low | 89 | 91 | 90 | 88 |
|
|
