Abstrato

The improved credit card customer behavior clustering analysis based on ant colony algorithm

Yingchun Liu

With the development of credit card, the banks need to classify the credit card customers with the advanced data mining technology, and then take different measures to target different customers. An improved K-means algorithm based on pheromone is proposed, which works with the transformation probability to realize the clustering and has reduced the number of the parameters and improved the speed of clustering. At last, the proposed algorithm is tested and used to analyze bank credit card customer spending behavior.