Two major performance bottlenecks with decision tree based classifiers in a data stream environment are the depth of the tree and the update overhead of maintaining leaf node statistics on an instance-wise basis to ensure ...