Automatic classification of uncertain data by soft classifier

Recently, the classification of uncertain data becomes a hot topic in data mining since more and more applications, such as sensor database, location database, biometric information systems, produce vague and imprecise data. Though there exist a lot of approaches to classify the uncertain data by hard classifiers, few of them address the classification of the uncertain data by soft classifier. In this paper, we propose an automatic soft classifier to classify the uncertain data. The automatic soft classifier first combines Fuzzy C-means with a fuzzy distance function to assign the uncertain data into their corresponding clusters. Then, the clusters are split automatically and incrementally based on an objective function until the value of the objective function reaches the threshold given by the user. The experiments show that automatic soft classifier works well in a database with uncertainties.