Abstract
The purpose of this paper is to provide an answer to the question of the relative effectiveness of the cosine function versus a polynomial function in the description and stability of prediction of a specific set of longitudinal data. If the data conforms to a known function (such as the cosine function), can we test for that function more effectively (that is to say by the stability of the weights upon cross-validation) than using a polynomial function for developing prediction equations?
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Copyright (c) 2002 Russel Brown, Isadore Newman (Author)
Downloads
Download data is not yet available.