MAFSIDS: a reinforcement learning-based intrusion detection model for multi-agent feature selection networks

Journal of Big Data

Table 8 Comparison of performance of multiple models in CSE-CIC-IDS2018 dataset

Detection Model	Accuracy	F1-Score	Test time(ms)
logistic regression	0.881	0.782	6.294
KNN	0.927	0.907	317.929
Random forest	0.836	0.735	26.870
GBM	0.934	0.921	24.717
Gaussian-NB	0.796	0.389	4.738
Bernoulli-NB	0.728	0.589	5.032
Multinomial-NB	0.558	0.498	4.496
AdaBoosts	0.946	0.906	59.217
Neural Network	0.901	0.809	30.525
XGBoost	0.947	0.933	130.918
DT	0.931	0.922	80.301
CNN-1D	0.929	0.918	98.374
DQN	0.941	0.925	110.327
DDQN	0.939	0.928	142.392
Our model	0.968	0.963	34.286