Skip to main content

Table 4 Classification results for the information theft dataset with sampling and tuned hyperparameters experiments; mean and standard deviations of AUC and AUPRC, (10 iterations of 5-fold cross-validation)

From: IoT information theft prediction using ensemble feature selection

Experiment name Mean SD Mean SD
AUC AUC AUPRC AUPRC
CatBoost Tuned RUS 1:3 0.99727 0.00188 0.99980 0.00032
CatBoost Tuned RUS 1:1 0.99803 0.00149 0.99979 0.00035
DT Tuned RUS 1:3 0.99362 0.00330 0.99524 0.00281
DT Tuned RUS 1:1 0.99357 0.00395 0.99337 0.00424
Light GBM Tuned RUS 1:3 0.99539 0.00305 0.99915 0.00106
Light GBM Tuned RUS 1:1 0.99569 0.00240 0.99925 0.00073
LR Tuned RUS 1:3 0.99495 0.00241 0.99872 0.00082
LR Tuned RUS 1:1 0.99499 0.00231 0.99658 0.01024
MLP Tuned RUS 1:3 0.99398 0.00372 0.99782 0.00188
MLP Tuned RUS 1:1 0.99420 0.00363 0.99710 0.00555
NB Tuned RUS 1:3 0.99266 0.00254 0.97693 0.00522
NB Tuned RUS 1:1 0.99247 0.00276 0.97723 0.00736
RF Tuned RUS 1:3 0.99671 0.00213 0.99960 0.00060
RF Tuned RUS 1:1 0.99719 0.00181 0.99967 0.00051
XGBoost Tuned RUS 1:3 0.99688 0.00198 0.99957 0.00071
XGBoost Tuned RUS 1:1 0.99711 0.00183 0.99957 0.00063