基于密度峰值聚类和局部稀疏度的过采样算法
|
吕佳, 郭铭
|
Oversampling algorithm based on density peaks clustering and local sparsity
|
Lü Jia, Ming Guo
|
|
表5 各算法在C4.5分类器上的各个指标评价结果
|
Table 5 Evaluation results of each index of each algorithm on the C4.5 classifier
|
|
Datasets | Evaluation | SMO | Bor | Safe | ADAS | SM⁃IPF | K⁃m S | ADAP | RSM | DPCLSO |
---|
D1 | F1⁃measure | 95.3504% | 95.9219% | 95.0244% | 95.9219% | 89.1739% | 94.6911% | 97.0330% | 95.9219% | 95.5800% | G⁃mean | 96.7189% | 97.0969% | 96.2903% | 97.0969% | 93.3878% | 97.0969% | 98.1526% | 97.0969% | 97.4750% | D2 | F1⁃measure | 85.2866% | 80.7499% | 85.6621% | 82.5853% | 88.1044% | 82.6621% | 83.5422% | 84.0379% | 86.7211% | G⁃mean | 90.6153% | 87.4349% | 89.4995% | 89.1953% | 92.7331% | 88.6294% | 88.9781% | 89.0365% | 91.4930% | D3 | F1⁃measure | 91.0761% | 90.7114% | 89.5812% | 89.7583% | 86.9262% | 89.2127% | 91.3965% | 90.4775% | 92.4308% | G⁃mean | 92.4931% | 92.2069% | 91.2686% | 91.5019% | 89.0857% | 90.9029% | 92.6818% | 92.0207% | 93.5368% | D4 | F1⁃measure | 58.0147% | 56.5508% | 55.8179% | 52.8216% | 58.5686% | 51.9370% | 58.2561% | 58.6345% | 57.5804% | G⁃mean | 62.2464% | 59.6573% | 61.0640% | 55.4117% | 43.2818% | 55.9819% | 63.7136% | 62.6257% | 63.3409% | D5 | F1⁃measure | 50.2477% | 46.2616% | 45.2641% | 47.9719% | 50.9139% | 49.9714% | 46.9857% | 45.5351% | 49.9561% | G⁃mean | 63.0937% | 59.8204% | 58.9365% | 61.2255% | 63.4165% | 62.8464% | 60.1788% | 59.0383% | 62.7733% | D6 | F1⁃measure | NaN | 38.3952% | NaN | NaN | 45.3598% | 35.6085% | 43.3759% | NaN | 46.5640% | G⁃mean | 56.1748% | 55.9361% | 44.1640% | 49.1841% | 63.9365% | 53.2240% | 59.0591% | 57.1104% | 63.5003% | D7 | F1⁃measure | 77.7990% | 76.7148% | 81.4483% | 75.0648% | 74.3187% | 77.9407% | 77.3053% | 79.2532% | 82.6409% | G⁃mean | 83.3495% | 82.6656% | 85.6686% | 81.4911% | 80.7767% | 83.1034% | 83.0608% | 84.4409% | 87.2673% | D8 | F1⁃measure | 40.8288% | 44.8296% | 42.0709% | 44.4995% | 52.4160% | 40.5528% | 33.3644% | 38.9963% | 47.1016% | G⁃mean | 55.3644% | 59.9155% | 56.7622% | 59.2967% | 65.0128% | 55.4519% | 49.6612% | 54.8542% | 60.7536% | D9 | F1⁃measure | 91.1449% | 90.9919% | 90.9771% | 91.0146% | 89.7471% | 91.1402% | 89.7448% | 90.8830% | 91.2878% | G⁃mean | 93.1435% | 92.9086% | 92.9363% | 92.9736% | 92.2880% | 93.1267% | 91.8047% | 92.9036% | 93.3944% | D10 | F1⁃measure | 76.2830% | 75.4806% | 76.8681% | 76.6538% | 77.1066% | 76.3623% | 76.6250% | 77.4209% | 77.7131% | G⁃mean | 82.7515% | 81.8017% | 82.8221% | 83.1245% | 84.2765% | 82.9522% | 82.9646% | 83.6477% | 83.8657% | D11 | F1⁃measure | 77.2319% | 77.2165% | 76.2536% | 77.0151% | 81.0328% | 78.3353% | 78.8932% | 79.1148% | 78.8529% | G⁃mean | 76.7661% | 77.3651% | 76.3400% | 77.0226% | 78.2896% | 77.9084% | 79.0216% | 79.1874% | 79.1526% | D12 | F1⁃measure | 96.4695% | 95.9850% | 96.8202% | 96.6032% | 94.0345% | 97.1781% | 96.0692% | 97.2658% | 97.5309% | G⁃mean | 98.1752% | 97.3734% | 98.0413% | 98.2613% | 96.4397% | 98.0484% | 97.7318% | 98.4311% | 98.6106% | D13 | F1⁃measure | 64.9351% | 69.5000% | 72.8778% | 71.1304% | 74.7826% | 69.3333% | 69.5652% | 65.3333% | 72.2826% | G⁃mean | 66.1703% | 72.3171% | 76.1999% | 73.9165% | 75.4439% | 72.2677% | 73.2411% | 69.1026% | 76.4439% | D14 | F1⁃measure | 92.0246% | 92.6147% | 91.7111% | 92.1847% | 82.0128% | 90.5300% | 90.1884% | 92.0944% | 93.2130% | G⁃mean | 96.0898% | 96.1157% | 95.7903% | 96.1208% | 90.4755% | 95.3021% | 96.1252% | 96.1801% | 96.2612% | D15 | F1⁃measure | 90.4892% | 60.4396% | 74.7253% | 79.1667% | 77.5000% | 90.4213% | 96.1538% | 89.0110% | 90.5983% | G⁃mean | 96.1528% | 71.2328% | 81.6052% | 85.5086% | 89.4334% | 96.2842% | 96.2910% | 89.5923% | 96.2328% | D16 | F1⁃measure | 76.4519% | 77.4593% | 75.9186% | 76.1207% | 74.6716% | 77.0081% | 77.1894% | 74.8543% | 76.6652% | G⁃mean | 88.7605% | 88.4716% | 86.6235% | 88.5958% | 82.3485% | 87.5854% | 90.1391% | 87.9138% | 88.2057% | D17 | F1⁃measure | 95.6210% | 95.8070% | 95.9727% | 94.0132% | 94.9890% | 96.0063% | 95.5166% | 95.4641% | 96.0063% | G⁃mean | 96.9293% | 96.8651% | 97.0851% | 95.6145% | 96.5988% | 96.9495% | 96.8329% | 96.6793% | 96.9495% | D18 | F1⁃measure | 95.5141% | 95.3489% | 95.2893% | 95.3475% | 94.5738% | 95.4256% | 94.9416% | 95.5958% | 95.8715% | G⁃mean | 95.6370% | 95.5051% | 95.4224% | 95.5040% | 94.6074% | 95.9596% | 95.1048% | 95.7345% | 96.0039% |
|
|
|