CMStatistics 2017: Start Registration
View Submission - CMStatistics
B1641
Title: Bayesian optimum warranty length under Type-II unified hybrid censoring scheme Authors:  Tanmay Sen - Indian Institute of Technology Patna (India) [presenting]
Yogesh Mani Tripathi - Indian Institue of Technology Patna (India)
Ritwik Bhattacharya - Indian Statistical Institute Kolkata (India)
Abstract: The determination of optimum warranty length is considered under Type-II unified hybrid censoring scheme. Consumers are willing to purchase a highly reliable product with certain cost constraint. To assure the product reliability and also to remain profitable, the manufacturer provides warranties on product lifetime. Moreover, censored lifetime data are available in practice, to assess the reliability of the product. Therefore, determination of an appropriate warranty length based on censored lifetime data is an important issue to the manufacturer. It is assumed that the lifetime follows a log-normal distribution. We consider a combine free replacement and pro-rata warranty policy (FRW/PRW). The life test is conducted under Type-II unified hybrid censoring scheme. The warranty length is obtained by maximizing an expected utility function.The expectation is taken with respect to the posterior predictive model for time to failure given the available data obtained under Type-II unified hybrid censoring scheme. A real data set is analyzed to illustrate the proposed methodology. We propose a non-linear prorate warranty policy and compare them with linear warranty policy. It is observed that non-linear prorate warranty policy give larger warranty length with maximum profit.