Multistart Algorithm for Global Optimization
نویسنده
چکیده
Abtstract A generalization of the multistart algorithm is proposed for nding the global minimizer of a nonlinear function of n variables. Our method concentrates a quasirandom sample by performing a few inexpensive local searches. The sample is then reduced by replacing worse points by new quasirandom points. A complete local search is performed only on those points with small function values. This method performs favorably in comparison to other global optimization methods.
منابع مشابه
A Derivative-Free Filter Driven Multistart Technique for Global Optimization
A stochastic global optimization method based on a multistart strategy and a derivative-free filter local search for general constrained optimization is presented and analyzed. In the local search procedure, approximate descent directions for the constraint violation or the objective function are used to progress towards the optimal solution. The algorithm is able to locate all the local minima...
متن کاملMultistart Local Search Continuous Global Optimization Method with a Taboo Step and its Condition for Finding the Global Optimum
We introduce a multistart local search-based method with a taboo step for solving continuous global optimization problems with bound constraints. Since this algorithm has a characteristic taboo step[5, 1995] by removing candidate points that converge to the current local optimum in each iteration, the step enables us to avoid repeated convergence to one of an already known optima in a local sea...
متن کاملSimulated Annealing for Sizing of Integrated Circuits in Spice
This paper presents a new optimization algorithm for automatic sizing of integrated circuits (IC) in SPICE. We refer to the new method as DESA. It is a hybrid between two very popular oprimization methods. The first one is differential evolution (DE) which is a robust population based optimization method and has received a lot of attention in the recent years. It was also sucessfully applied to...
متن کاملMultiple Solutions of Mixed Variable Optimization by Multistart Hooke and Jeeves Filter Method
In this study, we propose a multistart method based on an extended version of the Hooke and Jeeves (HJ) algorithm for computing multiple solutions of mixed variable optimization problems. The inequality and equality constraints of the problem are handled by a filter set methodology. The basic ideas present in the HJ algorithm, namely the exploratory and pattern moves, are extended to consider t...
متن کاملGlobal Optimization Strategies for Two-Mode Clustering∗
Two-mode clustering is a relatively new form of clustering that clusters both rows and columns of a data matrix. To do so, a criterion similar to k -means is optimized. However, it is still unclear which optimization method should be used to perform two-mode clustering, as various methods may lead to non-global optima. This paper reviews and compares several optimization methods for two-mode cl...
متن کامل