Table 2 Dataset size reduction using CSMCR.

From: Optimizing IoT intrusion detection with cosine similarity based dataset balancing and hybrid deep learning

Dataset

Dataset eecords

Balanced dataset size

Reduction in size

IoTID2041

352,516

77,196

78.10%

N-BaIoT42

498,164

124,308

75.05%

RT-IoT202243

123,117

25,014

79.68%

UNSW Bot-IoT44

3,999,055

13,490

99.66%