REGULARITY ESTIMATING METHOD FOR PATTERNS OF STREAMING DATA

摘要

Disclosed is a pattern regularity estimating method of streaming data. The method may be configured to comprise: an order group generating step of generating an order group according to the number of data set for order comparison of data; a streaming data input step of inputting analysis target streaming data in a buffer; a re-sampling step of removing noises from the streaming data and then performing a sampling operation as many as the number of data; a re-discretizing step of re-discretizing the re-sampled streaming data; a number allocation step for each size of comparing the sizes of the re-discretized data and allocating the number in the order of size; a belonging group determining and updating step of determining a belonging group in accordance with an allocated order and updating the number of members of the belonging group; and a pattern regularity calculating step of calculating pattern regularity based on the size of the data belonging to each group.