“Dimension Reduction: Feature Subset” Method for Selecting the Best Index Combination in Reputation Evaluation of Crowdsourcing Participants

Table 6

Evaluation results of crowdsourcing participants' reputation: evaluation index selection classifier confusion matrix.

Data dimensionality reduction method	Machine learning technology	Number of indicators	Accuracy	Precision rate			Recall rate			F1-measure
				P1	P2	P3	R1	R2	R3	F11	F12	F13
ReliefF	DT	9	0.9055	0.964	0.431	0.736	0.950	0.477	0.799	0.927	0.624	0.849
	SVM	9	0.9147	0.991	0.291	0.675	0.934	0.561	0.857	0.924	0.693	0.885
	BPNN	12	0.9132	0.987	0.321	0.672	0.938	0.488	0.838	0.925	0.611	0.873
	RBFNN	9	0.8987	0.980	0.363	0.555	0.936	0.432	0.761	0.917	0.560	0.822
	KNN	8	0.8958	0.957	0.375	0.729	0.942	0.412	0.815	0.919	0.561	0.853
	NB	13	0.8890	0.951	0.534	0.515	0.953	0.416	0.756	0.920	0.567	0.814