Fig. 4
Machine learning classification. Comparison of different machine learning models and their prediction capability in mapping the ionic current-residence time landscape. Each colored region represents an amino acid type (each letter indicates the type of amino acid) predicted by the models using the training data. Note that the colors in each plot are not correlated to the other ones. The solid dots along with the labels on the plots represent the mean values of the actual data (presented in Fig. 3a). a Prediction based on Nearest Neighbor Model with k = 3 b Prediction based on Logistic Regression (large red region belongs to the class of R). c Prediction based on Random Forest Model with the number of estimators = 9
