Dynamic fuzzy c-means (dFCM) clustering for continuously varying data environments

Many real world applications require online analysis of streaming data, making an adaptive clustering technique desirable. Most adaptive variations of available clustering techniques are application-specific, and do not apply to the applications of clustering as a whole. With this in mind, a generalized algorithm is proposed which is a modification of the fuzzy c-means clustering technique, so that dynamic data environments in differing fields can be addressed and analyzed. We demonstrate the capabilities of the dynamic fuzzy c-means (dFCM) algorithm with the aid of synthetic data sets, and discuss a possible application of the dFCM algorithm in associative memories, through preliminary experiments.