Bibliometric Analysis of the Application of Projection Pursuit Regression Method in Water Research in China

Abstract

【Background】 Practical problems in water science and engineering are diverse and complex, characterized by nonlinearity and multi-dimensions. Satisfactory solutions to them are not easy to obtain using traditional data analysis methods. The projection pursuit regression proposed in the 1960s can fill this gap and it has been developed and applied rapidly. Since its introduction to China in the 1980s, the projection pursuit method has seen a widespread applications because of its superiority in processing multi-dimensional data and flexibility in problem-oriented solving. 【Objective】 The objective of this paper is to review and summarize the evolution and application of the projection pursuit method in different disciplines in China since the 1980s, as well as the areas to which it has been mostly used. 【Method】 Our analysis was based on the literatures collected from the China National Knowledge Infrastructure database and considered four categories in the collection: temporal change, the disciplines, journals that published papers using this method, as well as some keywords representative of its applications. The data was then analyzed using the visualization analysis function of the bibliometrics provided by VOSviewer. 【Result】 Application and research of the projection pursuit method in China can be roughly divided into four stages: Introduction stage from 1984 to 1992; slow development stage from 1993 to 1999, rapid development stage from 2000 to 2010, and stable development after 2011. Its research and application are primarily in water conservancy engineering, environment science and engineering, and quantitative economy. Four of the top five journals that have published most projection pursuit papers are all water science journals, indicating that this method might be most suitable for solving water-related problems and are widely appreciated by the water science community. The areas that used the projection pursuit most included comprehensive evaluation, regression analysis, scheme selection and dynamic clustering. The results of clustering analysis showed that the research and application of the projection pursuit method in water-related problems are mainly in theoretical development, water quality evaluation, analyzing carrying capacity of water resource, water safety, and vulnerability of water resource systems. 【Conclusion】 The projection pursuit method has been widely used in China, especially in water science for evaluating water quality, water safety analysis, as well as bearing capacity of water resources. The method still has potential. In theoretical perspective, further work might focus on the fundamentals, improving the algorithms to determine the projection index function and the window width radius, as well as coupling it with other methods. In practical application, the method has potential applications in areas such as diagnosis identification, early warning, decision making, and evaluation of dynamic systems.

Keywords

• projection pursuit regression
• research progress
• water problem
• bibliometric method
• visualization analysis
• vosviewer