In this paper we propose a dimension reduced weighted fuzzy clustering algorithm (sWFCM-HD). The algorithm can be used for high dimensional datasets having streaming behavior. Such datasets can be found in the area of sensor networks, data originated from web click stream and data collected by internet traffic flow etc. These data’s have two special properties which separate them from other datasets: a) They have streaming behavior and b) They have higher dimensions. Optimized fuzzy clustering algorithm has already been proposed for datasets having streaming behavior or higher dimensions. But as per our information, nobody has proposed any optimized fuzzy clustering algorithm for data sets having both the properties, i.e., data sets with higher dimension and also continuously arriving streaming behavior. Experimental analysis shows that our proposed algorithm (sWFCM-HD) improves performance in terms of memory consumption as well as execution time Keywords-K-Means, Fuzzy C-Means, Weighted Fuzzy C-Means, Dimension Reduction, Clustering.
展开▼
