Search

Recently, controllability of complex networks has attracted enormous attention in various fields of science and engineering. How to optimize structural controllability has also become a significant issue. Previous studies have shown that an appropriate directional assignment can improve structural controllability; however, the evolution of the structural controllability of complex networks under attacks and cascading has always been ignored. To address this problem, this study proposes a new edge orientation method (NEOM) based on residual degree that changes the link direction while conserving topology and directionality. By comparing the results with those of previous methods in two random graph models and several realistic networks, our proposed approach is demonstrated to be an effective and competitive method for improving the structural controllability of complex networks. Moreover, numerical simulations show that our method is near-optimal in optimizing structural controllability. Strikingly, compared to the original network, our method maintains the structural controllability of the network under attacks and cascading, indicating that the NEOM can also enhance the robustness of controllability of networks. These results alter the view of the nature of controllability in complex networks, change the understanding of structural controllability and affect the design of network models to control such networks. [ABSTRACT FROM AUTHOR]

The controlling of complex networks is one of the most challenging problems in modern network science. Accordingly, the required energy cost of control is a fundamental and significant problem. In this paper, we present the theoretical analysis and numerical simulations to study the controllability of complex networks from the energy perspective. First, by combining theoretical derivation and numerical simulations, we obtain lower bounds of the maximal and minimal energy costs for an arbitrary normal network, which are related to the eigenvalues of the state transition matrix. Second, we deduce that controlling unstable normal networks is easier than controlling stable normal networks with the same size. Third, we demonstrate a tradeoff between the control energy and the average degree (or the maximum degree) of an arbitrary undirected network. Fourth, numerical simulations show that the energy cost is negatively correlated with the degree of nodes. Moreover, the combinations of control nodes with the greater sum of degree need less energy to implement complete control. Finally, we propose a multi-objective optimization model to obtain the control strategy, which not only ensures the fewer control nodes but also guarantees the less energy cost of control. [ABSTRACT FROM AUTHOR]

Differential evolution (DE) has been used to solve real-parameter optimization problems with nonlinear and multimodal functions for more than a decade of years. However, it is pointed out that this classical DE harbors restricted efficiency and limited local search ability. Inspired by that gradient-based algorithms have powerful local search ability, we propose a new algorithm, which is diversity-maintained DE based on gradient local search (namely, DMGBDE), by incorporating approximate gradient-based algorithms into the DE search while maintaining the diversity of the population. The primary novelties of the proposed DMGBDE are the following: (1) the gradient-based algorithm is embedded into DE in a different manner and (2) a diversity-maintained mutation is introduced to slow down the learning procedure from the searched best individual. We conduct numerical experiments with a number of benchmark problems to measure the performance of the proposed DMGBDE. Simulation results show that the proposed DMGBDE outperforms classical DE and variant without gradient local search or diversity-based mutation. Moreover, comparison with some other recently reported approaches indicates that our proposed DMGBDE is rather competitive. [ABSTRACT FROM AUTHOR]

For the global optimization problems with continuous variables, evolutionary algorithms (EAs) are often used to find the approximate solutions. The number of generations for an EA to find the approximate solutions, called the first hitting time, is an important index to measure the performance of the EA. However, calculating the first hitting time is still difficult in theory. This paper proposes some new drift conditions that are used to estimate the upper bound of the first hitting times of EAs for finding the approximate solutions. Two case studies are given to show how to apply these conditions to estimate the first hitting times. [ABSTRACT FROM AUTHOR]

Abstract: Robustness, the ability to maintain performance in the face of perturbations and uncertainty, is believed to be a necessary property of biological systems. In this paper, we address the issue of robustness in an important signal transduction network—epidermal growth factor receptor (EGFR) network. First, we analyze the robustness in the EGFR signaling network using all rate constants against the Gauss variation which was described as “the reference parameter set” in the previous study [Kholodenko, B.N., Demin, O.V., Moehren, G., Hoek, J.B., 1999. Quantification of short term signaling by the epidermal growth factor receptor. J. Biol. Chem. 274, 30169–30181]. The simulation results show that signal time, signal duration and signal amplitude of the EGRR signaling network are relatively not robust against the simultaneous variation of the reference parameter set. Second, robustness is quantified using some statistical quantities. Finally, a multi-objective evolutionary algorithm (MOEA) is presented to search reaction rate constants which optimize the robustness of network and compared with the NSGA-II, which is a representation of a class of modern multi-objective evolutionary algorithms. Our simulation results demonstrate that signal time, signal duration and signal amplitude of the four key components – the most downstream variable in each of the pathways: R–Sh–G–S, R–PLP, R–G–S and the phosphorylated receptor RP in EGRR signaling network for the optimized parameter sets have better robustness than those for the reference parameter set and the NSGA-II. These results can provide valuable insight into experimental designs and the dynamics of the signal-response relationship between the dimerized and activated EGFR and the activation of downstream proteins. [Copyright &y& Elsevier]

In this article, we propose a fast annealing genetic algorithm (FAGA), based on the principle of the minimal free energy in statistical physics, for solving multi-objective optimization problems. The novelties of FAGA are: (1) providing a new fitness assignment strategy by combining Pareto-dominance relation and Gibbs entropy, (2) introducing a new criterion for selection of new individuals to maintain the diversity of the population. We make many experiments to measure the performance of the proposed FAGA, and estimate its convergence rate for a number of test problems. Simulation results show that the FAGA is a very fast and effective algorithm. [ABSTRACT FROM AUTHOR]

Inference of the biochemical systems (BSs) via experimental data is important for understanding how biochemical components in vivo interact with each other. However, it is not a trivial task because BSs usually function with complex and nonlinear dynamics. As a popular ordinary equation (ODE) model, the S-System describes the dynamical properties of BSs by incorporating the power rule of biochemical reactions but behaves as a challenge because it has a lot of parameters to be confirmed. This work is dedicated to proposing a general method for inference of S-Systems by experimental data, using a biobjective optimization (BOO) model and a specially mixed-variable multiobjective evolutionary algorithm (mv-MOEA). Regarding that BSs are sparse in common sense, we introduce binary variables indicating network connections to eliminate the difficulty of threshold presetting and take data fitting error and the L0-norm as two objectives to be minimized in the BOO model. Then, a selection procedure that automatically runs tradeoff between two objectives is employed to choose final inference results from the obtained nondominated solutions of the mv-MOEA. Inference results of the investigated networks demonstrate that our method can identify their dynamical properties well, although the automatic selection procedure sometimes ignores some weak connections in BSs. [ABSTRACT FROM AUTHOR]