Open Conference Systems, ITACOSM 2019 - Survey and Data Science

Font Size: 
On influence of clustering population on accuracy of vector mean estimator
Grzegorz Antoni Sitek, Janusz Leszek Wywiał

Building: Learning Center Morgagni
Room: Aula Magna 327
Date: 2019-06-06 03:30 PM – 04:40 PM
Last modified: 2019-05-23


Vector of population means of variables under study is estimated on the basis of data observed in a sample drawn according several cluster designs.  Data on auxiliary variables are observed in the whole population. The variance-covariance matrices of vector estimators are depend on intra-cluster spread of the variables. The spread depends on a partition of the finite and fixed population into clusters characterized by matrix  of the intra-cluster homogeneity coefficients. The population is partitioned into clusters on the basis of the auxiliary variables. Accuracy of vector estimators are compared under several variants of clustering algorithms or sampling designs involving those dependent on auxiliary variables.

Full Text: SLIDES