A grid-based multistage algorithm for parameter simulation-optimization of complex system

Evolutionary algorithms (EA) are recently used to explore the parameter space of complex system simulations as the methodology sees models as black boxes. The first advantage is that these algorithms become independent from what kind of simulation has to be explored. The task is finding the parameter settings to optimize a given objective function. This optimization process evaluates the performance of possible parameter sets and converges towards the best alternatives. The evaluation needs to launch hundreds of thousands of simulation runs. This procedure copes with the combinatorial explosion of computation time and requires considerable computational resources. Furthermore, the original algorithms cannot guarantee the exploration in the search space uniformly and equally because the search is probabilistic. This work elaborates a multistage optimization process in a grid-enabled modeling and simulation platform. We propose a hybrid integration of various continuous optimization algorithms and optimize them for running with different Distributed Resource Management (DRM) systems. The performance of algorithm is compared to original algorithm in the optimization of Ants model.

[1]  Guillaume Deffuant,et al.  SimExplorer: Programming Experimental Designs on Models and Managing Quality of Modelling Process , 2010, Int. J. Agric. Environ. Inf. Syst..

[2]  Steven F. Railsback,et al.  Agent-Based and Individual-Based Modeling: A Practical Introduction , 2011 .

[3]  Riccardo Poli,et al.  Parallel genetic algorithm taxonomy , 1999, 1999 Third International Conference on Knowledge-Based Intelligent Information Engineering Systems. Proceedings (Cat. No.99TH8410).

[4]  Fabien Michel,et al.  EPIS: A Grid Platform to Ease and Optimize Multi-agent Simulators Running , 2011, PAAMS.

[5]  Fred W. Glover,et al.  Unidimensional Search for Solving Continuous High-Dimensional Optimization Problems , 2009, 2009 Ninth International Conference on Intelligent Systems Design and Applications.

[6]  James Smith,et al.  A tutorial for competent memetic algorithms: model, taxonomy, and design issues , 2005, IEEE Transactions on Evolutionary Computation.

[7]  David Meignan,et al.  Coalition-based metaheuristic: a self-adaptive metaheuristic using reinforcement learning and mimetism , 2010, J. Heuristics.

[8]  Julio R. Banga,et al.  Improved scatter search for the global optimization of computationally expensive dynamic models , 2009, J. Glob. Optim..

[9]  F. Azadivar.  Simulation optimization methodologies , 1999, WSC'99. 1999 Winter Simulation Conference Proceedings. 'Simulation - A Bridge to the Future' (Cat. No.99CH37038).

[10]  Forrest Stonedahl,et al.  Finding Forms of Flocking: Evolutionary Search in ABM Parameter-Spaces , 2010, MABS.

[11]  Shahryar Rahnamayan,et al.  A novel population initialization method for accelerating evolutionary algorithms , 2007, Comput. Math. Appl..

[12]  El-Ghazali Talbi,et al.  COSEARCH: A Parallel Cooperative Metaheuristic , 2006, J. Math. Model. Algorithms.

[13]  Richard J. Beckman,et al.  A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output From a Computer Code , 2000, Technometrics.

[14]  Sigurdur Olafsson,et al.  Chapter 21 Metaheuristics , 2006, Simulation.

[15]  Satoru Hiwa,et al.  Hybrid optimization using DIRECT, GA, and SQP for global exploration , 2007, 2007 IEEE Congress on Evolutionary Computation.

[16]  Uri Wilensky,et al.  NetLogo: A simple environment for modeling complexity , 2014 .

[17]  Guillaume Hutzler,et al.  Automatic Tuning of Agent-Based Models Using Genetic Algorithms , 2005, MABS.

[18]  David R. C. Hill,et al.  Declarative task delegation in OpenMOLE , 2010, 2010 International Conference on High Performance Computing & Simulation.