Research On Grey Cluster Of Panel Data And Its Applications

Grey clustering analysis is one of the important components of Grey system. It is widely used ineconomy, environment quality evaluation, military affairs, biological classifying, traffic problem, andso on. Traditional grey clustering method sets the analyzing model for the cross-section-data. It is lackof the research about grey clustering for panel data. According to the present researches, this thesisdoes further study on grey clustering for panel data. And then, apply this grey method in analyzingsome social and economic problems. The mainly research contents are described as following:(1) On the base of the traditional grey clustering, the conception of Grey time clusteringcoefficient has been introduced. This coefficient synthesizes the grey clustering coefficient matrixesin different time points to get the grey synthetical clustering coefficient matrix. So that brings out themethod of grey clustering for panel data. The method uses time coefficient in representation of timefactor. And it is used both in grey incidence clustering and grey whitenization weight functionclustering.(2) It is hard to confirm the whitenization weight function, when there is lack of subjectiveexperience or classify information about clustering objects. According to the definition of quartile instatistics, calculate the critical points of the whitenization weight function. Meanwhile, thewhitenization weight function can be set.(3) The method based on maximizing deviations has been improved to ensure the clusteringweight of every index. Avoid the complexity of the calculation of absolute value.(4) On basis of the qualitative analysis, the method of grey clustering for panel data is used toevaluate the developing status of eight sub provincial cities in economy, energy sources andenvironment. Finally, get the sorts of different developing status. And this proves the practicabilityand validity.