Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/87410
Authors: 
Exterkate, Peter
Year of Publication: 
2011
Series/Report no.: 
Tinbergen Institute Discussion Paper 11-138/4
Abstract: 
Kernel ridge regression is gaining popularity as a data-rich nonlinear forecasting tool, which is applicable in many different contexts. This paper investigates the influence of the choice of kernel and the setting of tuning parameters on forecast accuracy. We review several popular kernels, including polynomial kernels, the Gaussian kernel, and the Sinc kernel. We interpret the latter two kernels in terms of their smoothing properties, and we relate the tuning parameters associated to all these kernels to smoothness measures of the prediction function and to the signal-to-noise ratio. Based on these interpretations, we provide guidelines for selecting the tuning parameters from small grids using cross-validation. A Monte Carlo study confirms the practical usefulness of these rules of thumb. Finally, the flexible and smooth functional forms provided by the Gaussian and Sinc kernels makes them widely applicable, and we recommend their use instead of the pop ular polynomial kernels in general settings, in which no information on the data-generating process is available.
Subjects: 
nonlinear forecasting
shrinkage estimation
kernel methods
high dimensionality
JEL: 
C51
C53
C63
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
File
Size
174.15 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.