Tax sources category management is an important part in the tax administration. An important part of tax sources category is efficiency and rationality. An improved fuzzy clustering method used in tax sources classification is presented in this paper. This algorithm solves the disadvantage of losing information generated by hard classification of traditional clustering methods. The intrinsic characteristics between individuals can be revealed from a large number of tax-related data. The problem of focused management, clear management objectives and optimize resource allocation can be well resolved after taxpayers classified into different clusters. The experimental result also shows that the new improved fuzzy C means clustering algorithm combining with Parzen window estimation can resolve the initial central issue in original algorithm and reduce the clustering iterations.
展开▼
