A New K-Means Clustering Algorithm for Customer Classification in Precision Marketing

K-means is wildly used in data mining and clustering for its powerful data clustering ability, but its inherent limitations affect its application fields and accuracy. Theoriginal K-means algorithm is improved and applied in customer clustering in precision marketing. Firstly, integrates K-means algorithm with particle swarm optimization according to analyzing the source of the K-means calculation limitations; Secondly, improves the improved algorithm in its operation time, convergence speed, global solution exploration ability successively and redesigns the calculation procedures; Finally applies it in customer classification in precision marketing and the experiment results shows that the new algorithm can increasecustomer clustering effectiveness, validity, accuracy and has satisfactory results in practice.