Performance of learning of different algortithms on synthetic data sets

Long time no see, if anyone is reading this blog. Few days ago, I found very interesting visualization of how different ML algorithms are performing on synthetic data sets. The study conatins Bayes, KNN, regression, rule- and tree-based, SVM models. No wonder, that SVM with well-tuned RBF kernel gamma and cost parameters is performing very good, the same for tree-based models.