A FUZZY CLUSTERING METHOD FOR MINING CUSTOMER PREFERENCES

作者：Chou Jyh Rong^*

来源：Actual Problems of Economics, 2011, (123): 405-417.

• customer preference
• data mining
• fuzzy clustering
• clustering algorithm
• cluster validation

This paper presents a novel clustering method for mining customer preferences. The proposed method is based on the hierarchical clustering algorithm of combining fuzzy relations with grey relational grades. In addition to the clustering algorithm, a cluster validation index is proposed to validate the clustering results as well as to determine the best number of clusters. The proposed method is applicable to dealing with clustering problems involving complex discrete data. It can be used as a tool to help companies to extract preference patterns from large amounts of customers' preference decisions.

• 出版日期2011