CLASSIFYING OF MEASUREMENTS

摘要

The present invention relates to a method for determining a class label of captured attributes, comprising the steps of capturing (S101) a data stream of attributes using a sensor means; modeling (S103) a probability distribution for the captured attributes; traversing (S103) a decision tree starting from a root node, in which the nodes of the decision tree define split points of the probability distribution and the leafs of the decision tree define a probability value of the class label between the split points; and determining (S104) the class label of the captured attributes on the basis of the probability values as given by the leafs and the integral of the probability distribution between the split points as given by the nodes.