Multivariate Log – Normal Distribution–多元对数正态分布.PDFVIP

Multivariate Log – Normal Distribution Tarmast, Ghasem Statistics Department, Faculty of Mathematical Science and Computer Shahid Chamran University, Ahwaz , Iran. E – Mail: gasem_Tarmast@hotmail com. 1-Introduction Multivariate distributions have important role in economics and statistics. Since variables in economics, psychology, and reliability are positive; multivariate positive distributions should be considered. Johnson, and Kotz (1972) cited some multivariate positive distributions such as multivariate Gamma, Beta, Pareto, F distribution which have very complicated forms and are not easy to use in Economical and Reliability investigations. Jeevavand (1997) drive the Bayesian estimate of P (X X ) in bivariate Pareto distribution. 1 2 Tong (1980) developed the area of probability inequalities in multivariate distributions. Tarmast (1997) used multivariate normal distribution to obtain bounds for the multiple integral P (BXC). Here, multivariate log - normal distribution is defined and its mean and covariance matrix are obtained and their estimates are calculated. The application of the multivariate log - normal distribution in reliability is mentioned. 2-Multivariate Log-Normal Distribution Let X= [X , X ,…., X ] be a p-component random vector having multivariate Normal distribution 1 2 p with mean and covariance matrix D =( d ) . Now we use the transformation Yi = exp ( Xi ) and define ij Y=[Y ,Y ,….,Y ]. The density of Y is multivariate log - normal distribution and has the following form: 1 2 p