Discovering the patterns of energy consumption, GDP, and CO2 emissions in China using the cluster method

This paper explores the effects of enforced Chinese policies regarding environmental matters at the provincial level. The paper proposes several features and explores datasets from the last four Five-Year Plans, which extensively covered the period from 1995 to 2015. As the datasets are multidimensional, dimensional reduction is applied with a control of information loss described by data variance preservation. A data-based clustering algorithm is applied to identify the similarities, and characteristics are explored in relation to the specific features. The use of this method makes it possible to identify the changes in the relevance of the features in the most recent Five-Year Plan, in which decarbonization is relevant instead of productivity. This method also makes it possible to identify specific efforts several provinces have made to introduce cleaner energy production. Following the analysis, it is concluded that the proposed features enable the proposed method to be considered a tool for analyzing the real-world effects of the adopted policies. (C) 2018 Elsevier Ltd. All rights reserved.