Eman Taha Hassan Mohammed

Robust estimation in simultaneous equations model / تقدير حصين فى نموذج المعادلات الآنية Eman Taha Hassan Mohammed ; Supervised Ahmed Hassen Youssef , Ahmed Amin Elsheikh - Cairo : Eman Taha Hassan Mohammed , 2012 - 130Leaves : charts ; 30cm

Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Statistics and Econometrics

Optimization of LS estimator requires the normality of error distribution , where there are no outliers in dependent variable . Dropping this assumption affects the LS estimator . LS estimator is influenced by outliers , which usually associate with non normal distributions for errors whether heavy tail , asymmetric distrbution , or contaminated distribution . Outliers and influential points especially attract or pull LS estimator towards them , such that LS estimator efficiency decreases and becomes more bias



Contaminated distributions Diagnostic methods Robust estimator