Confusion Matrix - Discrimination from Molecular Dynamics trajectories

<p>One has 10 classes (the 9 endecapeptides + the control). The predictions were done on&nbsp;each data vector using the Random Forest classifier.&nbsp;</p> <p>Rows are the true labels and columns the predicted ones. The values are normalized by the total number of tests.</p> <p>First row would be ARG and of all predictions of the data 66% are true and 34% wrong distributed among the other amino acids as indicated in the cells.</p> <p>Colors are scaled from red (0) to blue (1) followig the values.</p>