加权多元经验过程的极限性质与两指标布朗桥的局部时-概率论与数理统计专业论文.docxVIP

AbstractStatistical Abstract Statistical inference theory is always based on the results of random sam- ple噩，恐， ，％of the population X．They are independent and identically distributed random elements with the same distribution as the population X． 馐 Generally,X is k-dimensional，and its distribution function _ ‘I． F(z)=P(x≤z)，z∈石产 is unknown．How to infer reasonably for the unknown distribution F is the most basic research subjects in the theory of non-parametric statistical inference． Usually,let R(z)= 去喜 1阢s盘1，p￡(z)=、／丽(jk(z)一F(z))，z∈R知， and call R as the empirical distribution of random sample and醴as the em- pirical process of random sample．The empirical process theory established by Glivenko-Cantelli theorem，Kolmogorov-Smirnov statistical theory,Donsker theo- rem etc．shows that the empirical distribution function and the empirical process are the most valid tool to solve it directly．The convergency of empirical processes are turned to be the core of the problem when the goodness-of-fit between the empirical distribution function and the true distribution function is considered， Furthermore，by the Skorohod theorem and Donsker theorems，it is well known that Brownian bridge is the limiting process of the empirical processes in some senses．With the development of research and its application，weighted empirical processes are gradually concerned by a lot of scholars over the last two decade． and research in this area have emerged from time to time． However，the existed results are more complete for the empirical distribution function and the empirical process in the case of k=1．When k≥2,that is，the research on the multivariate empirical process and multi-parameter Brownian bridge have many areas to be perfect．The paper attempts to do some researches on the local times of two-parameter Brownian bridge and a weighted multivariate III empirical empirical process．The local times of two-parameter Brownian bridge and the 嬲ylnptotic properties of the weighted empirical processe