Cosine-grey Correlation Analysis Model Based on Combination Weighting and Its Application

Abstract

Aiming at the problem that the traditional gray correlation model evaluation method has a single weight and the correlation degree is only related to distance, based on the existing improved methods, an evaluation method based on combined weighting-improved gray correlation is proposed. In order to increase the accuracy and credibility of the gray correlation, the proposed algorithm combines the improved methods of previous studies to dynamically determine the resolution coefficient, uses the improved entropy weight method and the coefficient of variation method to objectively weight the combination, combines traditional gray correlation analysis and cosine. Sorting method, comprehensively using multiple improvement methods to improve the gray correlation method. A case study of the correlation analysis of Wuxi's GDP and industry structure confirms that improving the model is feasible and effective.