Skip to main content

Table 3 The obtained results for MAUC from boosting algorithms using decision tree as a base learner

From: Boosting methods for multi-class imbalanced data classification: an experimental review

Dataset Algorithm
AdaBoost.MH SAMME CatBoost LogitBoost GradientBoost XGBoost MEBoost SMOTEBoost RUSBoost LightGBM AdaC1 AdaC2 AdaC3 AdaCost
Wine 0.9034 0.9139 0.9783 0.9574 0.9917 0.9659 0.9838 0.9499 0.9621 0.9693 0.9294 0.9164 0.9176 0.932
± 0.0183 ± 0.0128 ± 0.0054 ± 0.0098 ± 0.0025 ± 0.0129 ± 0.0077 ± 0.0167 ± 0.0091 ± 0.0152 ± 0.0168 ± 0.0162 ± 0.0134 ± 0.0139
Hayes-Roth 0.8064 0.7951 0.7588 0.8025 0.8032 0.7979 0.7941 0.7956 0.7734 0.5339 0.8013 0.8107 0.8162 0.8095
± 0.0155 ± 0.0104 ± 0.0154 ± 0.0087 ± 0.0138 ± 0.0151 ± 0.1901 ± 0.0153 ± 0.0292 ± 0.0343 ± 0.0107 ± 0.0115 ± 0.0197 ± 0.0173
Contraceptive 0.5281 0.4944 0.5305 0.4904 0.4939 0.5366 0.5032 0.4801 0.4458 0.5005 0.5455 0.3479 0.4422 0.5524
± 0.0048 ± 0.0071 ± 0.0046 ± 0.0061 ± 0.0100 ± 0.0072 ± 0.0051 ± 0.0121 ± 0.0074 ± 0.0060 ± 0.0056 ± 0.0044 ± 0.0114 ± 0.0093
Pen-based 0.9012 0.9581 0.962 0.9554 0.9513 0.9327 0.9726 0.9503 0.9592 0.9526 0.9035 0.9038 0.9029 0.9031
± 0.0055 ± 0.0055 ± 0.0045 ± 0.0053 ± 0.0030 ± 0.0057 ± 0.0037 ± 0.0027 ± 0.0047 ± 0.0020 ± 0.0096 ± 0.0056 ± 0.0061 ± 0.0079
Vertebral column 0.7758 0.7758 0.7999 0.7814 0.7794 0.7824 0.7814 0.8069 0.7851 0.7957 0.793 0.7973 0.7911 0.7881
± 0.0218 ± 0.0218 ± 0.0230 ± 0.0165 ± 0.0162 ± 0.0155 ± 0.0184 ± 0.0218 ± 0.0169 ± 0.0278 ± 0.0089 ± 0.0135 ± 0.0119 ± 0.0145
New thyroid 0.8879 0.8713 0.8899 0.9098 0.9063 0.8894 0.6225 0.9071 0.884 0.9097 0.8987 0.8728 0.8855 0.8833
± 0.0217 ± 0.0269 ± 0.0126 ± 0.0206 ± 0.0117 ± 0.0273 ± 0.0217 ± 0.0354 ± 0.0210 ± 0.0109 ± 0.0296 ± 0.0415 ± 0.0431 ± 0.0351
Dermatology 0.9551 0.9657 0.9783 0.9681 0.9715 0.9700 0.9187 0.9656 0.9438 0.9024 0.9558 0.9592 0.9517 0.9526
± 0.0103 ± 0.0050 ± 0.0033 ± 0.0087 ± 0.0057 ± 0.0123 ± 0.01691 ± 0.0062 ± 0.0113 ± 0.0121 ± 0.0107 ± 0.0087 ± 0.0101 ± 0.0108
Balance scale 0.6131 0.628 0.6719 0.6079 0.6103 0.6204 0.6268 0.6071 0.6618 0.7116 0.3603 0.6086 0.566 0.6893
± 0.0118 ± 0.0180 ± 0.0246 ± 0.0099 ± 0.0150 ± 0.0091 ± 0.0085 ± 0.0100 ± 0.0209 ± 0.0179 ± 0.0086 ± 0.0064 ± 0.0194 ± 0.014
Glass 0.6381 0.6644 0.6857 0.7187 0.6455 0.7087 0.6235 0.6662 0.5334 0.4205 0.6407 0.6479 0.5928 0.6412
± 0.0248 ± 0.0324 ± 0.0244 ± 0.0229 ± 0.0485 ± 0.0246 ± 0.0388 ± 0.0192 ± 0.0437 ± 0.0042 ± 0.0324 ± 0.0183 ± 0.0356 ± 0.0169
Heart 0.2839 0.2887 0.3054 0.2673 0.2838 0.2783 0.272 0.3216 0.2949 0.3182 0.2853 0.2781 0.2923 0.298
± 0.0136 ± 0.0112 ± 0.0131 ± 0.0147 ± 0.0171 ± 0.0181 ± 0.0136 ± 0.0202 ± 0.0144 ± 0.0389 ± 0.0152 ± 0.0257 ± 0.0145 ± 0.0165
Car evaluation 0.9848 0.9528 0.985 0.998 0.9851 0.8471 0.9909 0.9589 0.8439 0.9843 0.7711 0.8967 0.9478 0.9698
± 0.0102 ± 0.0115 ± 0.0034 ± 0.0025 ± 0.0075 ± 0.0098 ± 0.0046 ± 0.0086 ± 0.0196 ± 0.0036 ± 0.0278 ± 0.0046 ± 0.0047 ± 0.0049
Thyroid 0.9817 0.9754 0.9889 0.9769 0.9765 0.987 0.9718 0.9867 0.9612 0.8899 0.6667 0.9905 0.9907 0.6667
± 0.0033 ± 0.0023 ± 0.0026 ± 0.0038 ± 0.0029 ± 0.0022 ± 0.0070 ± 0.0021 ± 0.0168 ± 0.0325 ± 0.0000 ± 0.0022 ± 0.0010 ± 0.0000
Yeast 0.5088 0.4092 0.4757 0.4958 0.4619 0.539 0.4397 0.4889 0.2946 0.4209 0.4107 0.4806 0.5055 0.4846
± 0.0270 ± 0.0098 ± 0.0139 ± 0.0174 ± 0.0200 ± 0.0159 ± 0.0099 ± 0.0116 ± 0.0133 ± 0.0063 ± 0.0161 ± 0.0153 ± 0.0237 ± 0.0237
Page blocks 0.8183 0.7858 0.8645 0.8131 0.3038 0.8238 0.8117 0.825 0.8322 0.5566 0.7457 0.7391 0.8895 0.7455
± 0.0149 ± 0.0198 ± 0.0095 ± 0.0173 ± 0.0462 ± 0.0213 ± 0.0188 ± 0.0194 ± 0.0185 ± 0.0274 ± 0.0293 ± 0.0090 ± 0.0058 ± 0.0074
Shuttle 0.9479 0.9734 0.9607 0.9608 0.2682 0.9714 0.8121 0.9863 0.9727 0.3947 0.9457 0.9502 0.9467 0.9508
± 0.0177 ± 0.0119 ± 0.0121 ± 0.0213 ± 0.0404 ± 0.0158 ± 0.0149 ± 0.0127 ± 0.0165 ± 0.0317 ± 0.0146 ± 0.0168 ± 0.0116 ± 0.0170
Average 0.76896 0.7634 0.789 0.7802 0.6954 0.7767 0.7416 0.7797 0.7432 0.684 0.7102 0.7466 0.7625 0.7511
  1. The best performance is shown in italic for each dataset