Table 2.

Hypothetical placement of 60 Cases and Controls in training and testing sets

	Node 8	Node 9	Node 10	Node 11	Node 12	Node 13	Node 14	Node 15
Overall	0:6	15:0	0:17	2:0	35:9	2:8	2:20	4:0
Training	0:5	10:0	0:11	2:0	23:6	1:5	1:13	3:0
Testing	0:1	5:0	0:6	0:0	12:3	1:3	1:7	1:0

Overall placement of 60 Cases and 60 Controls (Cases:Controls) in the eight terminal nodes of the best decision tree shown in Table 1 and a hypothetical distribution between a training set and a testing set. Features 198, 140, 99, 68, 41, 95, and 251 are used in decision nodes 1 through 7 (Figure 1), respectively.