Year of Graduation
Investigation of Algorithms for Detecting the Structure of Significantly Overlapping Implicit Communities in Social Networks
Cluster analysis is an important part of data science. This paper focus on review and implementation of the recently invented algorithm EgoLP and its subprograms: egomunities retrieval, SLPA procedure, and LOBPCG. EgoLP and LOBPCG for searching for extremal eigenvalues of high-dimensional symmetric matrices are significant methods in their fields of science. The existence of the algorithms in open source C++ libraries meet needs of researchers who want to quickly and qualitatively extract information about the structure of communities from large graphs.