A note on the choice of the smoothing parameter in the kernel dinsity estimate
Daniel F. Froelich, Mezbahur Rahman · BRAC University Institutional Repository (BRAC University) · 2009
Among different density estimation procedures, the kernel density estimation has attracted the most attention. In this paper, the choices for smoothing parameter is discussed when the widely used Gaussian kernel is used in implementing the kernel density estimate. A simulation study is conducted from several mixtures of normal distributions covering a wide range of distributional shapes.