Accession Number : ADA115346

Title :   An Estimation Problem with Poisson Processes.

Descriptive Note : Technical rept.,

Corporate Author : STANFORD UNIV CA DEPT OF STATISTICS

Personal Author(s) : Gelfand,Alan E

PDF Url : ADA115346

Report Date : 20 Apr 1982

Pagination or Media Count : 31

Abstract : This paper considers an estimation problem involving n independent Poisson processes such that the i-th process has intensity function lambda (intensity)(t) = delta (intensity) p(t; alpha). It is of interest to estimate p(t; alpha). Two estimation procedures are developed, one using the exact arrival times of observations, the second using categorical arrival times of observations. Two specific instances of p(t), an exponential and a bilinear form are investigated further. An example applying the methodology to the active life of a judicial opinion is described. (Author)

Descriptors :   *Poisson density functions, *Estimates, Maximum likelihood estimation, Corrections, Observation, Time intervals, Arrival

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE