Flexible Selection of Wavelet Coefficients for Continuous Data Stream Reduction

In this article, we introduce a continuous data stream reduction method using wavelets summarization. Especially we consider storing a plenty of past data stream into stable storage (flash memory or micro HDD) rather than keeping only recent streaming data allowable in memory, because data stream mining and tracking of past data stream are often required. In the general method using wavelets, a specific amount of streaming data from a sensor is periodically compressed into fixed size and the fixed amount of compressed data (selected wavelet coefficients) is stored into stable storage. However, our method flexibly adjusts the number of selected wavelet coefficients for each local time section. Experimental results with some real world data show that our flexible approach has lower estimation error than the general fixed approach.