Study on Mutual Information Based Clustering Algorithm
Traditional clustering algorithms are designed for isolated datasets. But in most cases, relationships among different datasets are always existed. So we must consider the actual circumstances from the cooperative aspects. A new collaborative model is proposed and based on this model a new cooperative clustering algorithm is presented. In theorem, the algorithm is proved to converge to the local minimum. Finally, experimental results demonstrate that the clustering structures obtained by new algorithm are different from those of conventional algorithms for the consideration of collaboration and the performances of these collaborative clustering algorithms can be much better than those traditional separated algorithms under the cooperating circumstances.
Cited References Fulltext