Panel data clustering analysis based on composite PCC: a parametric approach

This paper proposed a panel data clustering model based on Hierarchical Nested Archimedean Copula (HNAC) model and compound PCC models. The method provides a new approach to panel data clustering, which breaks through the limitations of the traditional data clustering and time series clustering. This article makes full use of the dependence structure between the sectional individuals, as well as the degree of correlation between time series data. The similar structure was constructed by HNAC and Pair Copula to reflect the change of the clustering results. The selection of Copula clusters is very flexible giving the clustering results more accurate, robust, and easily interpreted. The computing efficiency is high and the estimation for the goodness-of-fit test are given based on compound PCC method in this paper. In the case study, the clustering results of compound PCC models are excellent. The result shows that the compound PCC models are effective and useful.