Fig. 1
From: Machine learning based characterization of high risk carriers of HTLV-1-associated myelopathy (HAM)
Illustration of the prediction results of anomaly carrier samples by the random forest classifier model. The left bar shows the training and test data for the classifier model. The model was trained on three sample groups and predicted the anomaly carrier samples as unseen test data. The anomaly carrier samples (n = 17) were classified into three prediction groups: Around 76.47% of the anomaly carrier data were predicted as HAM, whereas only 17.64% and 5.88% of the samples were predicted as carrier (n = 3) and ATL (n = 1) respectively (shown on the right bar).
