Premium
Measuring information in right‐censored models
Author(s) -
Hollander Myles,
Proschan Frank,
Sconing James
Publication year - 1987
Publication title -
naval research logistics (nrl)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.665
H-Index - 68
eISSN - 1520-6750
pISSN - 0894-069X
DOI - 10.1002/1520-6750(198710)34:5<669::aid-nav3220340507>3.0.co;2-l
Subject(s) - censoring (clinical trials) , majorization , variance (accounting) , mathematics , random variable , statistics , simple (philosophy) , variable (mathematics) , econometrics , computer science , discrete mathematics , accounting , mathematical analysis , philosophy , epistemology , business
The theory of majorization is used to develop measures of information in the randomly right‐censored model when the lifetime and censoring variables are discrete. Majorization also enables us to prove some basic theorems about the measures in a simple and unified fashion. These measures include Shannon's information as a special case. The measures are inadequate in the continuous case and some alternative measures, based on the variance of the lifetime random variable, are proposed.