Spectral Clustering: Advanced Clustering Techniques

Clustering is one of the widely using data mining technique that is used to place data elements into allied groups of “similar behavior”. The conventional clustering algorithm called K-Means algorithm has some well-known problems, i.e., it does not work properly on clusters with not well defined centers, it is difficult to choose the number of clusters to construct different initial centers can lead to different resultant clusters. Now a days,spectral clustering is becoming popular and widely used since its results overcomes the outcomes of the kmeans algorithm. Spectral clustering is a more advanced clustering algorithm compared to k-means as it uses several mathematical concepts (i.e. weight matrices,similarity matrices,degree matrices,similarity graphs,graph Laplacians ,eigenvalues and eigenvectors) in order to divide similar data points in the same group and dissimilar data points in