ESTIMATOR NADARAYA-WATSON DENGAN KERNEL ORDE BERHINGGA DAN TAK HINGGA

Maria Suci Apriani

Abstract


This article compared the performance between finite order kernel (normal) and infinite order kernel (sinus and cosinus) in Nadaraya Watson estimator by comparing the MSE values. This research used literature study method based on the article entitled Minimally Biased Nonparametric Regression and Autoregression by Timothy and Dimitris. The estimation result based on MSE values, showed that three kernels have different strengths on estimation process.


Keywords : Nonparametric regression, Nadaraya Watson estimator, finite order kernel, infinite order kernel.


Full Text:

PDF

References


Berg, Ar thur. 2008. Nonparametric Function Estimation with Infinite-Orde Kernels. Department of Statistics, University of Florida.

Hardle, Wolfgang. 1991. Smoothing Techniques With Implementation in S. Springer-Verlag, New York.

Hardle, Wolfgang. 1994. Applied Nonparametric Regression. Berlin

McMurry, T.L and Politis, D.N. 2003. Nonparametric Regression with Infinite Orde Flat-Top Kernel. www.math.ucsd.edu/~politis/PAPER/McMurryPolitis04.pdf, diakses pada tanggal 13 Februari 2013.

Wand, M.P and Jones, M.C. 1995. Kernel Smoothing. Chapman and Hall. London.


Refbacks

  • There are currently no refbacks.