An Efficient Approach to Detecting the Outliers in Forest Cover Data Streams

An important challenge in network management and intrusion detection is the problem of data stream classification. Concept-evolution is the open research issue in these circumstances, which involves the occurrence of a new class in the data stream. Most traditional data stream classification techniques are based on the assumption that the number of classes does not change over time. However, that is not the case in real world networks, and existing methods generally do not have the ability of identifying the evolution of a new class in the data stream. In this paper, we present an efficient approach for the detection of novel classes in data streams that exhibit concept-evolution. In particular, our approach is able to improve both accuracy and computational efficiency by eliminating “noise� clusters and false alarm rate in the analysis of concept evolution. Through an evaluation on simulated data sets, named as forest cover type, we express that our approach achieves comparable accuracy to an existing scheme from the literature with a significant decrease in computational complexity.