Classification of Customer Decision for Bicycle Purchase by Using the Ensemble Technique

Abstract

The objective of this research is to compare the efficiency of mean estimator that mean estimator which have been modified from the Horvitz-Thompson under adaptive cluster sampling when using simple Random unit selection. The simple estimation, ratio estimator and Rao-Blackwell estimator are used for estimation. The mean square error (MSE) is used as a criterion for comparison. The studied areas have been divided into 100 units. The populations are simulated by Poisson cluster process with parameter .Simulation of populations in 4 cases and sample size of 3 ,6, 9, 12, 16, 20 and 24. In each situation, data are simulated and repeated 1000 time. The result shows Rao-Blackwell estimator is most efficiency in 4 cases and in every sample size.