Regularized Projection Matrix Approximation with Applications to Community Detection

A. Subsequently, a clustering algorithm such as k-means or the In practical scenarios for community detection, the cluster EM algorithm is applied to identify the clusters. The success of information is typically not accessible beforehand. The affinity this method depends on the quality of the data representation matrix A is often computed using a kernel function or a cosine and the accuracy of the computational methods used for A. similarity function, which may deviate from the ideal assignment A popular approach for cluster identification is to utilize matrix. Persisting in applying the spectral projection the top K eigenvectors of matrix A, as employed in spectral approximation to derive the optimal rank-K projection matrix clustering [1], [2]. Identifying these eigenvectors is equivalent, approximation can result in a matrix X with negative elements.