Figure 3From: A meta-learning approach for B-cell conformational epitope predictionThree-level stacking architecture. The conformational epitope predictors and linear epitope predictors were all placed at Level 0. We selected C4.5, k-NN, and ANN as the Level-1 meta learners that transformed the output of the base predictors into meta features, and passed them to the successive level. We designated SVM as the top meta learner that learned from the base features and the meta features to produce the meta classification as the final result.Back to article page