A new method to evaluate whether the data are suitable to GM model or not

Abstract

Purpose

The purpose of this paper is to introduce the new class ratio dispersion, the new smooth degree sequence and the comparison criterion of the new smooth degree and to propose the new prior check of grey modeling in order to meet the modeling demand of the optimized grey models which have the white exponential law of coincidence.

Design/methodology/approach

This paper introduces the corresponding new concepts and new comparison criterion which can reflect the approach degree of the raw data and the normal geometric progression by analogy with the traditional class ratio dispersion, smooth degree sequence and comparison criterion.

Findings

To the optimized grey models, the new concepts and the new comparison criterion can be regarded as the prior check of grey modeling.

Originality/value

First, the new concepts and the new comparison criterion can reflect the approach degree of the raw data and the normal geometric progression, and this paper proposes the prior check of grey modeling to the optimized grey models. Second, this paper proposes the quantificational valuation criterion – the concept of the smooth degree which can reflect the approach degree of a single sequence and the normal geometric progression, and ends the status quo that there is only the comparison criterion of the smooth degree between two sequences but not the quantificational valuation criterion of a single sequence.

Citation

Wei, Y. and Hu, D. (2009), "A new method to evaluate whether the data are suitable to GM model or not", Kybernetes, Vol. 38 No. 7/8, pp. 1257-1264. https://doi.org/10.1108/03684920910976952