Classifier

The classifier algorithm allows rapid comparison of a diagnostic case with over 2.800 cases of a reference cohort and is bases on the random forest algorithm. The main output is a classification score that indicates the resemblance to one of the included central nervous system tumor classes. A score between 0 and 1 is generated for every class. All class prediction scores theoretically add up to 1. For a prediction, a methylation class score has to be above the cutoff of 0.9.