2020/05/06 by Shaik Tanveer Ul Huq, Huq, Shaik Tanveer ul, Vadlamani Ravi +3 · 1 citation
Computer Science · Physics and Astronomy · #97P80 #Complex Network Analysis Techniques #Data Visualization and Analytics #FOS: Computer and information sciences #I.2 #Neural and Evolutionary Computing (cs.NE) #Social and Information Networks (cs.SI)
paper · pdf · doi:10.48550/arxiv.2005.03181
openalex publication_date 2020/05/06 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
Most optimization-based community detection approaches formulate the problem\nin a single or bi-objective framework. In this paper, we propose two variants\nof a three-objective formulation using a customized non-dominated sorting\ngenetic algorithm III (NSGA-III) to find community structures in a network. In\nthe first variant, named NSGA-III-KRM, we considered Kernel k means, Ratio cut,\nand Modularity, as the three objectives, whereas the second variant, named\nNSGA-III-CCM, considers Community score, Community fitness and Modularity, as\nthree objective functions. Experiments are conducted on four benchmark network\ndatasets. Comparison with state-of-the-art approaches along with\ndecomposition-based multi-objective evolutionary algorithm variants (MOEA/D-KRM\nand MOEA/D-CCM) indicates that the proposed variants yield comparable or better\nresults. This is particularly significant because the addition of the third\nobjective does not worsen the results of the other two objectives. We also\npropose a simple method to rank the Pareto solutions so obtained by proposing a\nnew measure, namely the ratio of the hyper-volume and inverted generational\ndistance (IGD). The higher the ratio, the better is the Pareto set. This\nstrategy is particularly useful in the absence of empirical attainment function\nin the multi-objective framework, where the number of objectives is more than\ntwo.\n