Survey on Dynamic Concept Drift

摘要

The role of information technology and the advancement of the cloud computing have increased the use of the data in everyday life. Data analyzation and data storage become a challenging task during the processing of this large data. In the online applications, the data stream varies with rapid speed and has a larger volume. The recurring concept drift in the data stream makes the classification process to be complicated. Various algorithms discussed in the previous research works have not effectively addressed the problem of detecting the recurring concept drift in the data. The selection of the high-performing classifier model is also a challenging research goal. This research introduces two classification models for classifying the data with the recurring concept drift in the real time environment.