A machine learning framework for genotyping the structural variations with copy number variant

BMC Medical Genomics

Table 4 Performance comparison on real datasets-2

	The proposed method	BP	G-Features+BP	SVM with OVO	G-Feature+SVM with OVO
1	0.8875	0.9929	0.9868	0.9279	0.9897
2	0.8917	0.9677	0.9122	0.8753	0.8910
3	0.9083	0.8270	0.6190	0.5653	0.7699
4	0.9082	0.9251	0.7983	0.7131	0.8018
5	0.9000	0.7870	0.6417	0.4540	0.6057
6	0.9167	0.7324	0.6566	0.4017	0.5884
7	0.7258	0.6752	0.5807	0.6455	0.6016
8	0.8917	0.7724	0.7513	0.6131	0.6302
9	0.9083	0.8513	0.9250	0.9633	0.9230
Average	0.8820	0.8367	0.7635	0.6843	0.7557
Range	0.1909	0.3177	0.4061	0.5616	0.4013

ISSN: 1755-8794