Table 5 Comparison of state of the Art approaches on CICIDS2018 dataset.

From: Bi directional sparse attention recurrent autoencoder based intrusion detection for VANET security with tuna swarm optimization

Ref

Methods

FS-Method

AUC

F1-score

37

CatBoost

–

0.951

0.934

DT

–

0.910

0.886

NB

–

0.952

0.927

RF

–

0.954

0.921

XGBoost

–

0.911

0.887

38

DKNN-Kronecker_Neural_Network

CRDO

–

0.95

39

MP-CVAE (VariationalAutoEncoder)

–

0.95

0.96

40

ICVAE-BSM (Variational_AutoEncoder)

BPSO

0.965

0.956

41

ID-RDRL (Deep Q learning)

RFE

0.982

0.947

37

MAFIDS (Multi-Agent Feature Selection-DQL)

MFS

0.987

0.971

–

Our proposed model (DNN-BSAR-AE-LSO)

DNN

0.991

0.986

Back to article page