Table 3.

AD/MCI classification comparison of the ensemble methods (Voting, Uniform and Learned) and the other methods (Best and Average) in terms of accuracy, sensitivity and specificity when the training percentage varies from 1/2 to 3/4. In this experiment, a multi-source data including MRI, PET, Proteomics and CSF with 569 subjects in total.

Accuracy	Training Size	Voting	Uniform	Learned	Best	Average
	50.0%	0.8183	0.8177	0.8291	0.8278	0.8095
	66.7%	0.8288	0.8269	0.8337	0.8335	0.8182
	75.0%	0.8419	0.8298	0.8401	0.8401	0.8231
Sensitivity	Training Size	Voting	Uniform	Learned	Best	Average
	50.0%	0.5877	0.1965	0.2857	0.4339	0.2365
	66.7%	0.5926	0.2251	0.3017	0.4424	0.2598
	75.0%	0.5954	0.2218	0.315	0.4514	0.2631
Specificity	Training Size	Voting	Uniform	Learned	Best	Average
	50.0%	0.884	0.9916	0.9818	0.9924	0.9701
	66.7%	0.8953	0.9926	0.9804	0.9923	0.9722
	75.0%	0.9088	0.994	0.9818	0.9946	0.9743