Figure 4
From: Estimation of model accuracy by a unique set of features and tree-based regressor
Benchmark performance of MESHI_consensus. The results of 345 leave-one-target-out experiments are summarized in three histograms: (A) RMSE; (B) LOSS, the difference between the quality of the top-ranking model and the quality of the best one. (C) LOSS5, the minimal difference between the qualities of 5 top-ranking models and the quality of the best one. Median (red) and mean (green) values are indicated by vertical lines.
