earticle

영어

Data mining is a complex systematic engineering. For complex systems that is influenced by multi indexes, the relevance, hierarchy and fuzziness among indexes pose great challenges to data mining of complex system. Therefore, the paper proposes a multi-index grey relational data mining model of complex system based on grey system theory. First, the paper constructs index sets and grey categories for data mining objectives, and, under the different grey categories, builds up grey classic domain and grey partial domain for the aforementioned index sets. Next, normalized processing is undertaken to unify scales of different kinds of insets, whose weight generating algorithm is also provided herein. The paper then establishes a grey relational coefficient model and a grey relational degree model whose value is used to obtain the grey categories of the data mining objectives. Finally, the paper verifies and expounds multi-index grey relational data mining model by applying it to real-life practice. The results show the effectiveness and feasibility of this model.