Table 5.

The comparison of positive predicted value (PPV) between CA-125 alone and all four proteins (CA-125, MIF, OPN, and PROL) when using the four classification techniques, namely K-Nearest Neighbor (KNN), Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM). The PPV measures the liability of a positive test result and is defined as the ratio of the number of condition samples correctly identified to the total number of positive test results. We highlighted the cell(s) with best PPV in each experiment. In general, using logistic regression and all four proteins yield the best results on most occasions.

ID	Training	Testing	CA-125					4 Proteins
			KNN	LR	RF	SVM	Mean	KNN	LR	RF	SVM	Mean
1	E1 vs. H1	E1 vs. H1	1	1	0.42	1	0.86	1	1	1	1	1.00
2	E1 vs. H1	E2 vs. H2	0.32	0.32	0.24	0.43	0.33	0.87	0.92	0.70	0.61	0.78
3	E1 vs. H1	C2 vs. H2	0.88	0.88	0.76	0.95	0.87	1	0.97	0.94	0.91	0.96
4	L1 vs. H1	L1 vs. H1	1	0.91	1	1	0.98	1	1	1	1	1.00
5	L1 vs. H1	L2 vs. H2	0.63	0.63	0.63	0.86	0.69	0.89	0.92	0.82	0.87	0.88
6	L1 vs. H1	C2 vs. H2	0.62	0.70	0.62	0.85	0.70	0.93	0.90	0.85	0.88	0.89
7	C1 vs. H1	E2 vs. H2	0.10	0.12	0.12	0.15	0.12	0.26	0.24	0.13	0.14	0.19
8	C1 vs. H1	L2 vs. H2	0.61	0.69	0.59	0.76	0.66	0.88	0.79	0.81	0.80	0.82
9	C1 vs. H1	C2 vs. H2	0.63	0.67	0.57	0.76	0.66	0.82	0.84	0.83	0.85	0.84
10	E2 vs. H2	E2 vs. H2	0.33	0.33	0.17	0.33	0.29	0.67	0.80	0.60	0.67	0.69
11	E2 vs. H2	E1 vs. H1	0.83	1	0.60	1	0.86	1	1	1	1	1.00
12	E2 vs. H2	C1 vs. H1	0.97	1	0.91	1	0.97	1	1	1	1	1.00
13	L2 vs. H2	L2 vs. H2	0.87	0.85	0.58	0.85	0.79	0.94	0.89	0.85	0.94	0.91
14	L2 vs. H2	L1 vs. H1	1	1	0.79	1	0.95	0.98	1	0.85	1	0.96
15	L2 vs. H2	C1 vs. H1	1	1	0.83	1	0.96	0.94	1	0.87	1	0.95
16	C2 vs. H2	E1 vs. H1	0.86	0.92	0.41	1	0.80	0.59	0.71	0.33	0.65	0.57
17	C2 vs. H2	L1 vs. H1	1	1	0.81	1	0.95	0.96	1	0.88	1	0.96
18	C2 vs. H2	C1 vs. H1	1	1	0.85	1	0.96	1	1	0.87	1	0.97