Applied estimate for conditional bivariate Pareto distribution

摘要

A bivariate conditional Pareto distribution is derived from the marginal and conditional Pareto distributions. Its joint properties are derived including its posterior and the prior distribution, methods for estimating its parameters are discussed. It is applied to a bivariate dataset of Fertilizer A and Fertilizer B, where the conditional dependence of the two variables is suspected. The result shows that the model is compatible with the data. It is proposed for application when there is a need to account for such a conditional structure in a bivariate model.