Review Paper on Concept Drifting Data Stream Mining
International Research Journal of Modernization in Engineering Technology and Science · 2024
The Data Stream in dynamic and emerging environment such as e-commerce, financial data analysis, sensor systems, social networking and many more fields, that possess distribution.The term concepts refer to the whole distribution of the problem in a certain point in time and hence the concept drift represents a change in distribution of the problem.Data Stream that constantly changes with time due to some hidden concepts that exhibit varying degree of drift, often the magnitude and the frequency of drifting concept are not known apriori, which is very difficult to handle, because of inadequacy of traditional techniques.However, with the advent of streaming data and long life classification system, it becomes clear that training an accurate, fast and light classifier for unpredictable, large and growing environment is very important and still open research problem.