Paper Details

Mainly its inherent advantages of handling huge volumes of stream data and model drifting by usinggrouping learning is a common tool for data stream classification. The primarily focused on building accurate ensemblemodels from stream data is given in previous study. The linear scan of a huge number of base classifiers in the ensembleduring prediction incurs important costs in response time. Practical for much real-world time critical data streamapplications, such as Web traffic stream monitoring, spam detection, and intrusion detection preventing by grouplearning. It is necessary to classify each stream evidence in a timely manner in these applications, data streams usuallyarrive at a speed of GB/second. To deal with this problem, we propose a novel Ensemble-tree (E-tree for short) indexingstructure to organize all base classifiers in an ensemble for fast calculation. On one hand, E-trees treat group as spatialdatabases and employ an R-tree like height-balanced structure to decrease the predictable prediction time from linear tosub-linear complexity. On the other hand, E-trees can be continuously integrating new classifiers and discardingoutdated ones, well adapting to new trends and patterns underneath data streams updated by automatically. Theoreticalanalysis and experiential studies on both artificial and real-world data streams show the presentation of our approach.

Priyanka Narayan Kamble, Prof. Sonali.A.Patil, “Extended R-Tree Indexing Structure for Ensemble Stream Data Classification”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 3, Issue 3, pp. 284-288, March 2016.