Research Article

Feature Selection Based on Cross-Correlation for the Intrusion Detection System

Table 8

Improvement percentage of the CCFS method over MIFS and CFS methods for DT, SVM, KNN, and ND classifiers on the CIC-IDS2017 dataset.

CIC-IDS2017 (78 features)
Classification methodCCFS improvement versus MIFS (%)CCFS improvement versus CFA (%)
AccuracyPrecisionRecallF1-scoreAccuracyPrecisionRecallF1-score

DT2.804.073.703.881.071.611.251.43
SVM2.853.273.293.281.151.690.671.18
KNN2.553.123.123.120.731.090.750.92
NB2.462.132.122.131.831.651.051.35