Bayesian Mixture Model for Features-Preservation Clustering

The paper introduces feature preservation clustering which can handle the problems of privacy preservation and distributed computing. First, the Bayesian Mixture Model (BMM) are stated and some terminologies are de-fined. Second, Variational approximation inference for BMM is stated in detail. Third, base on the variational approximation inference, we design a distributed and paralleled algorithm for features preservation clustering. Finally, some datasets from UCI are chosen for experiment, Compared with K-means, the results show BMM algorithm does work better and BMM can work distributed and parallelled, so BMM can protect privacy information more and can save time.