Comprehensive evaluations of individual discrimination, kinship analysis, genetic relationship exploration and biogeographic origin prediction in Chinese Dongxiang group by a 60-plex DIP panel

Table 2 The confusion matrixes and statistics of the RF, XGBoost, SVM and DT models for the three continental (African, European and East Asian) biogeographic origin inferences, respectively

Random forest (RF)				Extreme gradient boosting (XGBoost)
Prediction of biogeographic origin	Africa	Europe	East_Asia	Prediction of biogeographic origin	Africa	Europe	East_Asia
Africa	165	2	0	Africa	162	1	0
Europe	0	122	0	Europe	3	123	2
East_Asia	0	1	703	East_Asia	0	1	701
Accuracy of biogeographic origin inference: 0.9970, 95% CI: (0.9912 ~ 0.9994)				Accuracy of biogeographic origin inference: 0.9930, 95% CI: (0.9855 ~ 0.9972)
Support vector machine (SVM)				Decision tree (DT)
Prediction of biogeographic origin	Africa	Europe	East_Asia	Prediction of biogeographic origin	Africa	Europe	East_Asia
Africa	163	3	1	Africa	148	10	4
Europe	1	120	3	Europe	14	103	9
East_Asia	1	2	699	East_Asia	3	12	690
Accuracy of biogeographic origin inference: 0.9889, 95% CI: (0.9803 ~ 0.9945)				Accuracy of biogeographic origin inference: 0.9476, 95% CI: (0.9319 ~ 0.9606)

ISSN: 1601-5223