On the Efficiency and Effectiveness of Controlled Random Search
- 349 Downloads
Applying evolutionary algorithms based on populations of trial points is attractive in many fields nowadays. Apart from the evolutionary analogy, profound analysis on their performance is lacking. In this paper, within a framework to study the behaviour of algorithms, an analysis is given on the performance of Controlled Random Search (CRS), a simple population based Global Optimization algorithm. The question is for which functions (cases) and which parameter settings the algorithm is effective and how the efficiency can be influenced. For this, several performance indicators are described. Analytical and experimental results on effectiveness and speed of convergence (Success Rate) of CRS are presented.
KeywordsControlled Random Search speed of convergence smooth optimization stochastic algorithms evolutionary algorithms effectiveness
Unable to display preview. Download preview PDF.
- Hendrix, E. M. T., Ortigosa, P. M. and García, I.: On Success Rates for controlled random search, Technical Note 00-01, Department of Mathematics Wageningen, 2000. (To appear in JOGO).Google Scholar
- Price, W. L.: A controlled random search procedure for global optimization, In: L. C. W. Dixon and G. P. Szegö (eds), Towards Global Optimization 2, North-Holland, Amsterdam, 1978, pp. 71–81.Google Scholar
- García, I., Ortigosa, P. M., Casado, L. G., Herman, G. T. and Matej, S.: Multidimensional optimization in image reconstruction from projections, In: I. M. Bomze, T. Csendes, R. Horst and P. M. Pardalos (eds), Developments in Global Optimization, Kluwer Acad. Publ., 1997, pp. 289–300.Google Scholar
- Nelder, J. A. and Mead, R.: A Simplex method for function minimization, Comput. J. 8 (1965), 308–313.Google Scholar
- Törn, A. and Zilinskas, A.: Global Optimization, Lecture Notes in Comput. Sci. 350, Springer-Verlag, Berlin, 1989.Google Scholar