Credit Risk Management of Consumer Finance Based on Big Data

Table 7

Comparison of the average performance evaluation index of each algorithm (rho = 0.7).

Rho = 0.7	Imbalance ratio	Accuracy	Recall rate	Precision	F1 measurement
SVM	10 : 1	0.8872	0.0735	0.5654	0.5289
Random forest		0.8737	0.2617	0.3891	0.5773
AdaBoost		0.9308	0.2918	0.7677	0.3942
Improve AdaBoost		0.8685	0.5841	0.5318	0.5306
SVM	20 : 1	0.9707	0.5960	1	0.6971
Random forest		0.9419	0.3153	0.1318	0.6891
AdaBoost		0.9693	0.4278	0.7277	0.6368
Improve AdaBoost		0.9178	0.6932	0.3934	0.5571