Author(s)

Long Wang

Corresponding Author

Long Wang

Abstract

The research on the application detection algorithm of merchant category code is carried out with data mining technology. Based on the data mining technology, the basic innovation work is carried out, and a new clustering algorithm – “fast hierarchical clustering algorithm based on root-seeking” (RS algorithm) is proposed. A comparative experiment is performed for the algorithm. Experimental results show that RS algorithm is superior to other classical algorithms in most data sets. Then a complete set of merchant category code application detection model is proposed. Five classification algorithms are selected and tested on four data sets. The experimental results show that the proposed merchant category code application detection model is feasible, and its accuracy rate and recall rate on three data sets are above 80%. To sum up, theoretical innovation is carried out with data mining technology, and a “merchant category code application detection model” is proposed based on the application scenario of data mining in the financial field. The effectiveness of the proposed algorithm and model is verified by a lot of experiments.

Keywords

Data Mining, Hierarchical Clustering, Merchant Category Code Application, Industry Model