Skip to main content
Fig. 1 | BioData Mining

Fig. 1

From: Gene function finding through cross-organism ensemble learning

Fig. 1

Process of the proposed ensemble method. The annotation matrices of the source (Homosapiens) and target (Musmusculus) organisms are extracted from the GPKB; with n (n=5 in the example) different perturbations on the former one, n matrices are created and used to train n different prediction models, each of them is then applied on the target annotation matrix; out of the single model predictions, a voting approach (3 out of 5 in the example) selects the final predicted annotations

Back to article page