Comparison Of The Estimators Of The General Least Squares Method Of The Parametric Model With The Estimators Of The Nadaria-Watson Weighted And Unweighted Method Of The Non-Parametric Model Of Longitudinal Data.

Abstract

Abatract

The water is polluted as a result of the presence of foreign substances, whether organic or inorganic. These substances spoil the water quality, which is what happens to the Tigris River. These waste products pollute the waters of the Tigris River, harmful to human health, causing diseases, and among these diseases is an amoeba, whose data represented by the numbers of people infected with the disease were taken on both sides of Karkh and Rusafa in Baghdad for the year 2018 as a dependent variable (Y) in a complete series (12) month. The concentrations of water pollutants were taken on the number of cases of the disease. Seven pollutant concentrations were relied upon, which represent the explanatory variables (Xi), which represent cross-sectional data for ten stations located on the banks of the Tigris River on both sides of Karkh and Rusafa. Because of dealing with cross-sections and time series, Panel data is the best name for it. We used parametric and non-parametric Panel data models with several non-parametric estimators, which are the Nadaria Watson weighted estimator and the Nadaria Watson unweighted estimator to determine which of the estimators is the most efficient that gives the best model, and the number of recorded injuries In hospitals and health centers on both sides of Karkh and Rusafa, which suffer from the problem of heterogeneity of variance, and that the variables of pollutant concentrations suffer from the problem of autocorrelation, and that the appropriate parametric model is the random effect model for the panel data according to the Houseman test.