Function (biology)

Algorithm

Smoothing

Biology

Statistics

Support vector machine

Selection (genetic algorithm)

Computer science

Mathematical optimization

Kernel smoother

Radial basis function kernel

Evolutionary biology

Kernel (algebra)

Regression

Mathematics

Kernel method

Nonparametric statistics

Nonparametric regression

Artificial intelligence

Combinatorics

Kernel regression

Regression function

The problem of deciding how much to smooth is of greatimportance in nonparametric regression. Before embarking ontechnical solutions of the problem it is worth noting that aselection of the smoothing parameter is always related to acertain interpretation of the smooth. However, a goodautomatically selected parameter is always a useful starting(view)point. An advantage of automatic selection of thebandwidth for kernel smoothers is that comparison betweenlaboratories can be made on the basis of a standardized method.Various methods for choosing the smoothing parameter arepresented in the following sections. The choice is made so thatsome global error criterion is minimized. This paper shortlyaspires to summarize attained results from this branch and todemonstrate their application for simulated data sets.

Kernel Estimation of the Regression Function - BandwidthSelection