Table 4 Performance analysis for dataset 1.

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

Metrics

CIC-IDS 2017 dataset

Proposed

AlexNet

DenseNet

SqueezeNet

GoogleNet

Sensitivity

0.985

0.950

0.973

0.954

0.953

Specificity

0.990

0.954

0.972

0.951

0.949

Accuracy

0.991

0.966

0.979

0.957

0.955

Precision

0.985

0.951

0.972

0.950

0.949

Recall

0.985

0.940

0.973

0.934

0.929

F-measure

0.985

0.943

0.974

0.943

0.941

NPV

0.991

0.932

0.975

0.951

0.949

FPR

0.011

0.023

0.033

0.034

0.035

FNR

0.037

0.125

0.145

0.146

0.147

MCC

0.985

0.955

0.972

0.951

0.950

Training Time (s)

320

400

560

340

450

Testing Time (s)

85

100

110

92

98

Inference Time (ms)

15

20

25

18

22

Back to article page