Search

Purpose: The purpose of this paper is to construct a grey incidence model for panel data that can reflect the incidence direction and degree between indicators. Design/methodology/approach: Firstly, this paper introduces the concept of a negative matrix and preprocesses the data of each indicator matrix to eliminate differences in dimensions and magnitudes between indicators. Then a model is constructed to measure the incidence direction and degree between indicators, and the properties of the model are studied. Finally, the model is applied to a practical problem. Findings: The grey-directed incidence degree is 1 if and only if corresponding elements between the feature indicator matrix and the factor indicator matrix have a positive linear relationship. This degree is −1 if and only if corresponding elements between the feature indicator matrix and the factor indicator matrix have a negative linear relationship. Practical implications: The example shows the number of days with good air quality is negatively correlated with the annual average concentration of each pollutant index. PM2.5, PM10 and O3 are the main pollutants affecting air quality in northern Henan. Originality/value: This paper introduces the negative matrix and constructs a model from the holistic perspective to measure the incidence direction and level between indicators. This model can effectively measure the incidence between the feature indicator and factor indicator by integrating information from the point, row, column and matrix. [ABSTRACT FROM AUTHOR]