IMPROVED ROUGH FUZZY POSSIBILISTIC C-MEANS (RFPCM) CLUSTERING ALGORITHM FOR MARKET DATA

Despite the wide variety of techniques available for grouping individuals into Market segments, K-means clustering algorithm is the popular and widely used method. The k means clustering algorithm aims to partitioning the given ‘n’ number of observations into k clusters to minimize an objective function. But the main shortcoming of k means clustering algorithm is its computational complexity and its performance in terms of execution time. In order to get the better performance Rough Fuzzy Possibilistic C-Means (RFPCM) algorithm is proposed. RFPCM algorithm increases the speed of grouping the clusters. This algorithm uses Rough C-Means (RCM), Fuzzy C-Means (FCM), and Possibilistic C-Means (PCM) algorithms. Further to improve the performance of RFPCM, the proposed method uses two different approaches: Ranking method and Query Redirection Method. Ranking Method facilitate to estimate the likelihood of the occurrence of data items or objects. It helps to create cluster that are having similar properties between all data point with in that cluster. The another approach is Query Redirection method which provides a mechanism to retrieve the data in the form of tables applicable to the request can be satisfied by more than one Logical table sources. RFPCM with Ranking Approach (RFPCMRA) and RFPCM with Query Redirection Approach (RFPCMQRA) are implemented in Matlab and the result shows that RFPCMQRA gives better performance compared with RFPCMRA in terms of execution time and computational complexity.