Issue No. 03 - May-June (2013 vol. 10)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TCBB.2013.72
Yang Tang , Humboldt University of Berlin, Berlin and Potsdam Institute for Climate Impact Research, Potsdam
Huijun Gao , Harbin Institute of Technology, Harbin
Jurgen Kurths , Potsdam Institute for Climate Impact Research, Potsdam, Humboldt University of Berlin, Berlin, and University of Aberdeen, Aberdeen
In this paper, we investigate the multiobjective identification of controlling areas in the neuronal network of a cat's brain by considering two measures of controllability simultaneously. By utilizing nondominated sorting mechanisms and composite differential evolution (CoDE), a reference-point-based nondominated sorting composite differential evolution (RP-NSCDE) is developed to tackle the multiobjective identification of controlling areas in the neuronal network. The proposed RP-NSCDE shows its promising performance in terms of accuracy and convergence speed, in comparison to nondominated sorting genetic algorithms II. The proposed method is also compared with other representative statistical methods in the complex network theory, single objective, and constraint optimization methods to illustrate its effectiveness and reliability. It is shown that there exists a tradeoff between minimizing two objectives, and therefore pareto fronts (PFs) can be plotted. The developed approaches and findings can also be applied to coordination control of various kinds of real-world complex networks including biological networks and social networks, and so on.
Biological neural networks, Optimization, Controllability, Complex networks, Vectors, Sorting, Evolutionary computation
Y. Tang, H. Gao and J. Kurths, "Multiobjective Identification of Controlling Areas in Neuronal Networks," in IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol. 10, no. 3, pp. 708-720, 2013.