Table 9 Performance analysis of the proposed model over SOTA methods of dataset 3.

From: A secure and efficient deep learning-based intrusion detection framework for the internet of vehicles

Metrics

CICIoV2024 dataset

Proposed

DFSENet [15]

GAN [18]

TCAN-IDS [22]

XAIEM [23]

Sensitivity

0.962

0.941

0.957

0.944

0.93

Specificity

0.986

0.977

0.965

0.968

0.95

Accuracy

0.976

0.963

0.955

0.946

0.94

Precision

0.98

0.97

0.95

0.965

0.96

Recall

0.968

0.955

0.966

0.953

0.945

F-measure

0.974

0.962

0.962

0.948

0.94

NPV

0.988

0.97

0.963

0.965

0.95

FPR

0.015

0.02

0.025

0.03

0.031

FNR

0.038

0.04

0.045

0.05

0.06

MCC

0.973

0.96

0.968

0.965

0.95

Training Time (s)

500

550

600

590

610

Testing Time (s)

100

110

120

115

125

Inference Time (ms)

12

15

18

16

17

Back to article page