A goodness-of-fit test for a class of bivariate Laplace distributions

摘要

A goodness-of-fit test for a class of bivariate Laplace distributions is proposed and studied. The distributions in this class are obtained as differences of two Moran-Downton bivariate exponential distributions. The resultant distributions have correlated univariate Laplace marginals. The class of distributions considered can be successfully used to model financial, biological or engineering data. To construct a goodness-of-fit test for this family, we consider a test statistic that takes advantage of the convenient formula of the characteristic function of this distribution and compares it with the empirical characteristic function of the sample. Large sample properties of the proposed test are studied. Some numerical results are reported which study the finite sample performance of the proposed test, as well as some real data set applications which illustrate the practical application of the proposed technique.