Decision Trees Constructing over Multiple Data Streams

Abstract

In this paper, we consider the decision trees-building question over multiple data streams. The online data mining over each data stream is often computationally prohibitive. Therefore, we propose an analysis for the problem of a large specimen. It concerns a case of satisfactorily large cardinality of samples. With the help of a given algorithm for the decision tree construction, we obtained data multistream decision trees. Experimental results demonstrate that the received decision trees can be efficient tools for the classification analysis.