Abstract
In this article, we develop a general statistical inference procedure for the probability of successful startup p in the case of startup demonstration tests when only the number of trials until termination of the experiment are observed. In particular, we define a class of startup demonstration tests and present expectation-maximization (EM) algorithm to get the maximum likelihood estimate of p for this class. Most of well-known startup testing procedures are involved in this class. Extension of the results to Markovian startups is also presented.
Acknowledgment
The author is grateful to the editor and two referees for their insightful comments that have led to improvement in an earlier version of the article.
Additional information
Notes on contributors
Serkan Eryilmaz
Prof. Serkan Eryilmaz is currently Vice President for research at Atilim University. He is the present co-chair of the System Reliability Technical Committee of the European Safety and Reliability Association. He served as Associate Editor for the IEEE Transactions on Reliability, and is currently serving as an Area Editor of the IISE Transactions-Quality & Reliability Engineering. Professor Eryilmaz’s research interests include reliability, applied probability and stochastic modeling. He is the author of more than 150 publications.