Fig. 3
From: Unsupervised encoding selection through ensemble pruning for biomedical classification
The box plot shows the Matthews correlation coefficients (MCC) distribution (y-axis) using a 100-fold Monte Carlo cross-validation. The rows depict the base with various encodings, and the columns depict the pruning method. The respective ensemble types can be found in the x-axis. The MVO pruning (mvo) has only been conducted on the first five folds (see Discussion). The MCC of the individual classifiers (single) has been collected for the best (Top_1), second (Top_2), and third (Top_3) model per fold, respectively. The plot shows the example of the avp_amppred dataset