Statistical inference on the accelerated competing failure model from the inverse weibull distribution under progressively type-II censored data

• Main Authors: Wang Ying, Yan Zai-Zai
• Format: Article
• Language: English
• Published: VINCA Institute of Nuclear Sciences 2021-01-01
• Series: Thermal Science
• Subjects: constant-stress accelerated competing failure model; thermal stress; inverse weibull distribution; markov chain monte-carlo method
• Online Access: http://www.doiserbia.nb.rs/img/doi/0354-9836/2021/0354-98362100097W.pdf
• ISSN: 0354-9836; 2334-7163

In this paper, the parameter estimation is discussed by using the maximum likelihood method when the available data have the form of progressively censored sample from a constant-stress accelerated competing failure model. Normal approximation and bootstrap confidence intervals for the unknown parameters are obtained and compared numerically. The simulation results show that bootstrap confidence intervals perform better than normal approximation. A thermal stress example is discussed.