Tolerance limits for inverse gaussian distribution

The inverse Gaussian distribution has been recognized as a versatile lifetime model with sound physical interpretation. However, it is not widely used as some of its important characteristics have not been obtained. In this paper, we derive simple expressions for one-sided lower tolerance limits of the inverse Gaussian distribution where the parameters are unknown. These proposed limits are constructed from confidence intervals of the parameters which are also available for multi-censored sample. A computationally simpler and less conservative approximation is also proposed. Monte Carlo simulations are carried out to evaluate these limits in terms of their average values and coverage probability.