AN EMPIRICAL SURVEY ON FREQUENT EXTRACTIONS IN DATA STREAMS OF MOBILE NETWORKS

Abstract

Frequent data mining patterns is a stimulating and difficult problem due to the emergence of recent applications and restricted resources in terms of main memory and time interval. Based on the lack of existing algorithms, this paper analyzes the sliding window model and proposes an effective sliding window data mining model algorithm. This method provides an extracted dynamic sliding window using a series of simple lists of existing items in the window. For every component of the window, chooses the most extreme stockpiling proficiency list type dependent on its recurrence to store the presence data. New window alteration methods that incorporate rundown type transformation are utilized to control memory use as the concept changes. Compared with the recently proposed algorithm, the results show that the prevalence of the proposed method is a lot of orders of magnitude in terms of memory usage and execution time.