| In the field of big data,temporal data are widely used in the Internet,the Internet of things,many aspects,such as sensors and scientific experiments,such as the daily stock price data,temperature,humidity,such as traffic time series data.It has the characteristics of large volume,diversity,low value density and fast speed.But as the era of information explosion produces huge amounts of time series data is growing,valuable data and no value has been the growth of the crazy,the value of the potential for such a large sequence data information is more difficult.As a result,information processing becomes more challenging.The purpose of this paper is to extract all the data information from the time series data.Due to the infinite growth of massive time series data,the storage scale of the expansion of the increase.Based on the data of any information to minimize data and information available to maximize information processing strategies,only stored data,the control data is not stored,not dominate the calculation of the value of data information with data,can be extracted effectively.In this paper,affine transformation model is adopted to realize the processing of time series data information.It mainly involves two core contents: first,controlling the acquisition of data.The arrival time series data is collected through clustering,and then the mean migration method and the least connected dominating set idea are respectively adopted to obtain the dominating set in each cluster,and finally the dominating set is stored in the storage structure.Second,data information extraction.In the process of acquiring the dominating data,according to the constraint of approximate results,it is calculated by the dominating set and affine matrix in the storage structure.Then,comparative experiments are carried out for the two methods to obtain the dominating set respectively,which proves that the mean migration method is superior to the connected dominating set in a certain sense. |
PDF Full Text Request