chapter11ClusAdvanced.ppt

* * * * * * * * * * But, how to compute the similarity efficiently? Computing inner product of two NxN matrices is too expensive. * Very expensive to compute directly We convert it into another form * * * * * * * * * * * * * * * * * * * * * * * * * * Summary Probability Model-Based Clustering Fuzzy clustering Probability-model-based clustering The EM algorithm Clustering High-Dimensional Data Subspace clustering: bi-clustering methods Dimensionality reduction: Spectral clustering Clustering Graphs and Network Data Graph clustering: min-cut vs. sparsest cut High-dimensional clustering methods