Author's: Mei Ling Huang
Pages: [61] - [78]
Received Date: March 8, 2010
Submitted by:
This paper studies estimation of the Pareto distribution. We use a weighted empirical distribution function (WEDF) for estimating the population distribution function of the Pareto distribution. The exact efficiency function of this estimator relative to classical empirical distribution function (EDF) is derived. Properties of this efficiency function are studied. Monte Carlo simulation results confirm the theoretical conclusions, which indicate that the WEDF is more efficient than the EDF in many situations. Application to kernel and parametric density estimation is also given.
efficiency, order statistics, weighted empirical distribution function.