Characterizations Based on Conditional Expectations of Order Statistics

• Main Authors: Tzu-Fang Kuo, 郭慈芳
• Other Authors: Wen-Jang Huang
• Format: Others
• Language: en_US
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/47735406524113069552

碩士 === 國立中山大學 === 應用數學系研究所 === 88 === It is known that record values and order statistics are closely related. When record values and order statistics are viewed as point processes, the two processes both share the order statistics property. The results of Beg and Balasubramanian(1990), Wu and Ouyang(1996), and Huang and Su(1999) about record values and order statistics motivated us to investigate more general results of characterization for order statistics point processes by using conditional expectations based on order statistics. On the other hand, in the class of point processes, there are a lot of characterizations of homogeneous Poisson processes based on the memoryless property of exponential distribution. The result of Asadi(1999) about characterization of the Gumble bivariate exponential or the bivariate geometric distribution inspired us be interested in investigating some similar results about non-independent bivarite homogeneous Poisson processes.