Optimal Data Streams Clustering Algorithm Based On N-δ Sliding Window Model

With the continuous upgrading of software and hardware and the rapid development of WEB technology,the information technology gains tremendous breakthrough,which leads the data collection into an automatic and intelligent process.At the same time,many fields have gained a mass of data sets.Besides this,the appearance of data streams introduces the further development and gains a widely application.For example,in the finance field and tele-communication field,the mass data could be abstracted as a new data model which could be called as data streams,which is not only a kind of opportunity but also a challenge.The rapid development of science and technology as well as the emergence of streaming data makes the research of streaming data mining technologies as the current hotspot.The application and development of streaming data mining technology is not only an opportunity but also a challenge,comparing with traditional data mining model.This paper mainly discusses the research of clustering algorithm based on the streaming data models.The paper has proposed an optimal clustering algorithm of streaming data, improved Clustream algorithms and data structure and expanded the sliding window model.By the theoretical and experimental analysis,the improving algorithm has gained an optimal algorithm complication and efficiency.This paper has done some jobs and gains results just as follows:First,this paper has defined the structure of algorithm based on computing cluster,incorporating and maintaining the OTCF structure and the process of clustering results.For one thing,OTCF structure,which introduces a parameter of t~*,is an optimal algorithm structure.At the same time,it gives an eliminated mechanism for higher efficiency of multidimensional streaming data,which could release some storing space and improve the using and dealing efficiency of algorithm.For another thing,the paper extends a new sliding window,whose size is changing with a certain time period.At the end,it gives an optimal clustering algorithm which named as OpClAlg.Based on this algorithm,we could deal streaming data model at real time and make it full use in many fields.Second,this paper gives an eliminated mechanism for higher efficiency of multidimensional streaming data,which could release some storing space and improve the using and dealing efficiency of algorithm.For another thing,the paper extends a new sliding window,whose size is changing with a certain time period.At the end,it gives an optimal clustering algorithm which named as OpClAlg.Based on this algorithm,we could deal streaming data model at real time and make it full use in more fields.Third,the paper has implemented the algorithms,which has used the programming designing language of Java and the database designing tool of SQL Server 2000 and is used in tele-communication query fields.It indicates that the optimal cluster streams algorithm,which based on extended N-δsliding window,has lowered the computing complication, improved the use factor of synopsis data structure and gained a better clustering quantity.Therefore,the new algorithm not only gains a quick computing velocity but also could be used effectively in many fields.