A Controlled Random Search Algorithm with Local Newton-type Search for Global Optimization
In this work we deal with the problem of finding an unconstrained global minimizer of a multivariate twice continuously differentiable function. In particular we propose an algorithm which combines a controlled random search procedure based on the modified Price algorithm described in  with a Newton-type unconstrained minimization algorithm proposed in . More in particular, we exploit the skill of the Price strategy to examine the whole region of interest in order to locate the subregions “more promising” to contain a global minimizer. Then starting from a point in these regions, we use an effective Newton-type algorithm to compute very quickly the closest local minimizer. In this way we succeed in improving the efficiency of the Price approach. Numerical results on as set of standard test problems are reported with the aim to put in evidence the improvement in efficiency when dealing with large scale problems.
Unable to display preview. Download preview PDF.
- 5.A. V. Levy and A. Montalvo. The tunneling algorithm for the global minimization of functions. SIAM Journal of Optimization Theory and Applications, 62 (2), 1989.Google Scholar
- 7.S. Lucidi, F. Rochetich, and M. Roma. Curvilinear stabilization techniques for truncated Newton methods in large scale unconstrained optimization. To appear on SIAM Journal on Optimization, 1998.Google Scholar
- 11.W.L. Price. A Controlled random search procedure for global optimization. In L.C.W. Dixon and G.P. Szego, editors, Towards Global Optimization 2. North-Holland, Amsterdam, 1978.Google Scholar