Skip to main content

Table 4 LDC for tabular datasets

From: D-Confidence: an active learning strategy to reduce label disclosure complexity in the presence of imbalanced class distributions

Dataset Classifier ff.ldc c.ldc dc.ldc Best
Iris SVM 65.3 6.7 2.7 dc
Iris NNET 65.3 83.0 1.3 dc
Iris RPART 65.3 3.3 1.7 dc
Cleveland SVM 24.2 39.8 14.5 dc
Cleveland NNET 24.2 16.2 10.5 dc
Cleveland RPART 24.2 17.7 13.2 dc
Poker SVM 200.4 250.9 98.2 dc
Poker NNET 200.4 311.6 114.2 dc
Poker RPART 200.4 326.5 168.4 dc
Satlog SVM 107.0 34.1 9.6 dc
Satlog NNET 107.0 24.4 15.9 dc
Satlog RPART 107.0 16.0 11.4 dc
Vowels SVM 88.6 31.3 26.3 dc
Vowels NNET 88.6 27.3 24.3 dc
Vowels RPART 88.6 31.6 39.9 c