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

Table 2

Total variance of original variables explained by principal components.

Principal component	Initial eigenvalue			Extract the sum of squares of loads
	Total	Variance contribution rate (%)	Cumulative percentage (%)	Feature value	Variance contribution rate (%)	Cumulativepercentage (%)
1	5.484	19.587	19.587	5.484	19.587	19.587
2	4.699	16.781	36.369	4.699	16.781	36.369
3	1.983	7.084	43.452	1.983	7.084	43.452
4	1.636	5.842	49.295	1.636	5.842	49.295
5	1.577	5.634	54.929	1.577	5.634	54.929
6	1.243	4.439	59.367	1.243	4.439	59.367
7	1.218	4.352	63.719	1.218	4.352	63.719
8	1.046	3.737	67.456	1.046	3.737	67.456
9	1.031	3.683	71.139	1.031	3.683	71.139
10	1.008	3.600	74.738	1.008	3.600	74.738
11	0.967	3.454	78.192
12	0.946	3.377	81.569
13	0.858	3.063	84.632
14	0.738	2.635	87.268
15	0.680	2.428	89.696
16	0.528	1.887	91.583
17	0.485	1.733	93.316
18	0.357	1.276	94.592
19	0.320	1.143	95.735
20	0.303	1.083	96.818

Note. List the principal components with weight ranking 1–20. The eigenvalues of the 10th principal components are greater than 1 in bold.