Exact likelihood inference for Laplace distribution based on Type-II censored samples
G. Iliopoulos and N. Balakrishnan
We develop exact inference for the location and scale parameters of the Laplace (double exponential) distribution based on their maximum likelihood estimators from a Type-II censored sample. Based on some pivotal quantities, exact confidence intervals and tests of hypotheses are constructed. Upon conditioning first on the number of observations that are below the population median, exact distributions of the pivotal quantities are expressed as mixtures of linear combinations and of ratios of linear combinations of standard exponential random variables, which facilitates the computation of quantiles of these pivotal quantities. Tables of quantiles are presented for the complete sample case.
Key words and phrases: Laplace (double exponential) distribution; exact inference; maximum likelihood estimators; Type-II censoring; mixtures; pivotal quantities; linear combinations of exponential order statistics.
Download (250Kb) (The original publication
is available at www.sciencedirect.com)