Fig. 3
From: VENturing into machine learning for the morphological analysis of von Economo neurons
Most important variables considering the classification algorithm. The importance of variables is represented by their median ranked order (x values and point numbers), where lower values indicate higher importance. Across all machine learning classifiers, average length emerged as the most significant variable. Specifically, it held the top spot in importance for C50, Earth, and XGBoost, and was the second most important variable for SVM. Inter-algorithm correlations for variable importance rankings ranged from r = 0.78 to r = 0.94 (all p < 0.001), demonstrating high consistency among machine learning methods. Average inter-algorithm correlation: r = 0.87 ± 0.05.
