Application of Projection Pursuit Dynamic Cluster Model in Regional Partition of Water Resources in China

The present study develops a projection pursuit dynamic cluster (PPDC) model by combining dynamic cluster with projection pursuit to solve the problem of regional partition of water resources in China. The procedures of the PPDC model are described as follows. Firstly, a multi-factor cluster problem can be converted into a single-factor (projected characteristic value) cluster problem according to linear projection. Secondly, a new projection index on the basis of dynamic cluster rule is set up in the PPDC model, which successfully avoids the problem of parameter calibration and makes objective cluster results. Thirdly, genetic algorithm (GA) is applied to optimize projection direction of the PPDC model. Finally, the developed PPDC model is used in a case study of regional partition of water resources in China to evaluate its application. The cluster results of the PPDC model agree well with the actual regional partition of water resources in China, indicating that the PPDC model is a powerful tool in multi-factor cluster analyses and could be a new method for regional partition of water resources. Copyright Springer Science+Business Media B.V. 2008