Abstract
In this paper we discuss the approximation of life distributions by exponential ones. The main results are: (1) ∀F∈ NBUE, where its mean is 1, we have\(|\bar F(t) - e^{ - t} | \leqslant 1 - e^{ - \sqrt {20} } ,\forall t \geqslant 0\), ∀≥0, where ρ = 1 - μ2/2, μ2 being the second moment ofF. The inequality is sharp. (2) In the case ofF∈IFR, the upper bound is\(1 - e^{ - \tfrac{\rho }{{1 - \rho }}} \). (3) For the HNBUE class, the upper bound is min\((\sqrt[3]{{4\rho }}.\sqrt[3]{{4\rho }})\). Furthermore, the improved upper bound is\(\sqrt[3]{{36\rho /(3 + 2\sqrt \rho )^2 }}\). In addition, we show\(\mathop {\sup }\limits_{t > 0} |\bar G(t) - e^{ - t} | \leqslant \sqrt {\frac{\rho }{2}} \), where\(\bar G(t) = \int_t^\infty {\bar F} (u)du\) (4) For the IMRL class, the upper bound is ρ/(1+ρ) ([1]). Here we give a simple proof.
Similar content being viewed by others
References
Brown, M., approximating IMRL Distributions by Exponential Distributions, with Applications to First Passage Times,Ann. Prob.,11 (1983), 419–427.
Brown, M. and Ge Guangping, Exponential Approximations for Two Classes of Aging Distributions, in China-Japan Symposium on Statistics, 1984, Beijing.
Барзилович, Е. Ю., и др., Вопросы математической теорий надеЗности, Москва, Радио и Связь, 1983.
Author information
Authors and Affiliations
Additional information
Project supported by the National Natural Science Fund of China.
Rights and permissions
About this article
Cite this article
Cheng, K., He, Z. Exponential approximations in the classes of life distributions. Acta Mathematicae Applicatae Sinica 4, 234–244 (1988). https://doi.org/10.1007/BF02006220
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02006220