两个点过程之间的相关系数的计算-概率论与数理统计专业论文.docxVIP

Point Point process is a very important stochastic process and it is a hot problem of the international academic frontier research．In the published literature，People often use a single number(correlation coefficient)to describe the correlation between the two point processes directly．In this article，we point out that the single number approch could not fully port- 一rait the correlation relationship between two point processes，and we USe correlation coefficient curve(CCC)to portrait the relationship between th一 -em，andvalidate that correlation coefficient curve can completely Portra- -it the correlation of two point processes in the neural network model． The main work of paper is below： The first part mainly introduces the history and the state of developm一 -ent of the point process，the research state of correlation coefficient，the the main conclusion and arrangement of this paper． In second part，some definition used in this article are description de一 -taily，for example，counting process，renewal process． The third part is the core of the article．It is di fferent from the traditi一 ．onal method：We use correlation coe筒cient curve to describe their corre． -lation relationship dynamically，and obtain a very important conclusion： n two two point processes，their correlation coefficient curve is a function that is dependent on binsize and increasing．We also confirm the range of correl一 ．ation coefficient． In the fourth part，conbining the correlation coefficient curve with t- 一he neural network model makes a testament．We apply our approach to data from a simple neural model Integrate—and—Fire model．It is found that the I-F model behaviors as a decorrelator(it reduces the positive input CO一 一rrelation coefficient，and increases negative correlation coefficient)． Key words：correlation coefficient；binsize；point process；renewal process；counting process． m 英文摘要 英文摘要 II 1绪论 1．1点过程的历史和发展现状 (1) 1．2相关系数研究现状 (3) 1．3主要结果及本文结构安排 ．．(4) 2基本知识 2．1计数过程 (7) 2．2平稳点过程 ．．(8) 2．3更新过程 (8) 3更新过程之