Table 2 Performance comparison of the proposed model against baseline architectures.

From: An enhanced deep learning framework for intrusion classification enterprise network using multi-branch CNN-attention architecture

Method

Accuracy (%)

Recall (%)

Specificity (%)

Precision (%)

F1-Score (%)

Training Time (hours)

Inference Time (ms)

Model Complexity (# Params)

1D-CNN

96.12

95.4

95.9

95.88

95.64

1.5

25

0.8 million

LSTM

96.88

96.01

96.65

96.2

96.1

2.3

35

1.5 million

CNN-LSTM

97.45

97.13

96.84

97.29

97.2

2.8

40

1.8 million

CNN + XGBoost

98.41

98.07

98.3

98.45

98.25

3.2

38

2.0 million

Proposed (CNN-Att + DT)

99.28

99.25

99.12

99.31

99.28

3.5

32

2.3 million

Back to article page