. Author manuscript; available in PMC: 2017 Oct 1.

Published in final edited form as: Neuroinformatics. 2016 Oct;14(4):439–452. doi: 10.1007/s12021-016-9307-8

Table 5.

performance comparisons of separate and joint association study by our proposed DGMM in terms of Root Mean Square Error (RMSE) and Correlation Coefficient (CC)

Method	RMSE(mean±std)		CC(mean±std)

	train	test	train	test
DGMM(VBM)	0.7886±0.0079	0.8214±0.0314	0.3688±0.0207	0.2484±0.0570
DGMM(FDG)	0.7901±0.0087	0.8229±0.0262	0.3605±0.0195	0.2345±0.0676
DGMM(AV45)	0.7877±0.0085	0.8201±0.0322	0.3805±0.0176	0.2545±0.0572
DGMM(A-ALL)	0.7841±0.0083	0.8164±0.0293	0.4309±0.0116	0.2895±0.0269
DGMM(W-All)	0.7838±0.0087	0.8160±0.0285	0.4393±0.0133	0.2938±0.0149