The fuzzy clustering on market segment

In any study of market segmentation, researchers often use clustering analysis as a tool. The analysis often is in a crisp partition form. But in practice, the sample are usually not well distributed, therefore the form may not be precisely defined. That is, one sample can belong to two or more groups. But, due to the fact that the requirements on the consumers and on the market are very high and the many real-market problems are fuzzy by nature and not random, the probability applications have not been very satisfactory in a lot of case. In this study, we adopt the fuzzy cluster method and attempt to combine a new compactness and separation validity function to build market segmentation in order to address the fuzziness among the group boundaries. Then we can use membership grade to describe each group. Therefore, the real market situation is clearly presented. Through membership grade, we depict the reality of the market, which lies between integers and real number. Buyers´ mindsets are both rational and complicated, so their purchasing decisions are not predictable and are affected by many factors. The structural stability of the market can be tested by the loyalty of buyers who pertain to different clusters. Marketing strategies will also have effects on the movement of group for housing buyer