Comparison of four clustering methods
Feng Xiao · Microcomputer & its Applications · 2010
The article described the more common four kinds of clustering algorithm:k-means,hierarchical clustering,SOM and FCM,described the principle and their use steps,and then verified and compared these algorithms using internationally accepted test data set IRIS.The results showed that for the type of the test data,FCM and k-means have a high level of accuracy,hierarchical clustering has the worst accuracy,while the SOM is the longest time-consuming.