Fuzzy clustering algorithmic and rationality test model
Shouyu Chen, Shuying Wang · 2004
The paper provides fuzzy clustering rule function to derive the optimal fuzzy cluster matrix and center matrix. Farther, two different analysis approaches, i.e., generalized Euclidean weighted distance analysis and weighted F-statistic test, are combined to provide new methods of rationality test. It analyses the weights of seven important indexes through integration of reasoning analysis and pairwise comparison method, and selects representative gauging stations in Minjiang basin by combining fuzzy clustering analysis, rationality test and main principle of setting up gauging stations. The application shows that the method is reasonable and efficient, and that the distance analysis and test are of significance.