شماره ركورد كنفرانس :
4109
عنوان مقاله :
On the asymptotic behaviour of the positive-rule Stein-type kernel density estimator
پديدآورندگان :
Mahmoodi M Department of Statistics, School of Mathematical Sciences,Shahrood University of Technology, Shahrood, Iran , Arashi M Department of Statistics, School of Mathematical Sciences,Shahrood University of Technology, Shahrood, Iran , Nezakati A Department of Statistics, School of Mathematical Sciences,Shahrood University of Technology, Shahrood, Iran
كليدواژه :
Asymptotic distributional mean square error , Bandwidth , Kernel density estimator , Positive , rule shrinkage estimator
عنوان كنفرانس :
يازدهمين سمينار ملي احتمال و فرآيندهاي تصادفي
چكيده فارسي :
When it is suspected that the true density of population may be f0(x), using shrinkage
strategies, one can improve kernel density estimator of the unknown f(x). In this paper, using
the prior information f0(x), with the aid of asymptotic theory we show how one can improve the
performance of kernel density estimator, using positive-rule Stein-type kernel density estimator.
Hence, the asymptotic performance of this estimator is compared with a strong competitor. From
theoretical results along with numerical computations, the superiority conditions of the newly
proposed estimator are investigated.
