Imputacja danych w spisach powszechnych
TL;DRAbstract
Results of a simulation experiment aimed at an appraisal of utility of some original model of data imputation in censuses are presented in the paper. It is based on clustering of records-donors according to their similarity and on the method of statistical roulette, i.e. a rotational algorithm arranging to records receives the lacking data in a random way from the nearest homogeneous donor clusters. The exercise, which showed high efficiency of the applied attempt, was performed using data for the gmina Gołuchów in the Wielkopolska region collected during the National Population and Housing Census conducted in 2002.
Chat with Paper
AI Agents for this Paper
Results of a simulation experiment aimed at an appraisal of utility of some original model of data imputation in censuses are presented in the paper. It is based on clustering of records-donors according to their similarity and on the method of statistical roulette, i.e. a rotational algorithm arranging to records receives the lacking data in a random way from the nearest homogeneous donor clusters. The exercise, which showed high efficiency of the applied attempt, was performed using data for the gmina Gołuchów in the Wielkopolska region collected during the National Population and Housing Census conducted in 2002.
Keywords
Chat
Click to start Chat