One-Step Dynamic Classifier Ensemble Model for Customer Value Segmentation with Missing Values

Table 7

McNemar’s test for pairwise comparison of performance in “China churn” dataset.

Model	LMF	KI-SVM	MS-SVM	EM-SVM	RI-SVM
ODCEM	15.254 (0.0000)	13.136 (0.0003)	12.971 (0.0003)	11.529 (0.0007)	14.469 (0.0001)
LMF		3.0480 (0.0809)	4.0364 (0.0367)	11.256 (0.0008)	0.0280 (0.8676)
KI-SVM			0.1880 (0.6650)	4.6880 (0.0304)	3.1840 (0.0744)
MS-SVM				2.7130 (0.1102)	3.5000 (0.0614)
EM-SVM					11.726 (0.0006)

Note: the results listed in Table 7 are the Chi squared values and P values are in brackets.