Kernel adjusted density estimation is a modification of the regular kernel density estimation. The modification is applied to a kernel function. This kernel function is derived from the location-scale transformation. Simulation study shows that this estimation have better results than the regular estimation because it has smaller MSE value. In addition, if normal kernel is used as a kernel function then the curve estimation will be smoother than other kernel function such as uniform kernel and Epachenikov kernel.
Copyrights © 2016