CLUSTERING METHODS IN GLOBAL OPTIMIZATION
暂无分享,去创建一个
Global optimization methods are designed to be used for problems where the objective function may have several local minima (maxima). Such problems are in general unsolvable and therefore the methods are designed to maximize the probability of discovering the global minimum. One class of such methods are the so called clustering methods. In these computations are made to concentrate sampled points around the local minima so that they can be recognized by a cluster analysis technique. Using these methods multiple determination of the local minima (multiple local optimizations) can be avoided and more work spent on the global exploration thereby increasing the probability that the global minimum will be found.
[1] Alexander H. G. Rinnooy Kan,et al. A stochastic method for global optimization , 1982, Math. Program..
[2] A. A. Torn. Cluster Analysis Using Seed Points and Density-Determined Hyperspheres as an Aid to Global Optimization , 1977, IEEE Transactions on Systems, Man, and Cybernetics.
[3] Aimo A. Törn,et al. A sampling-search-clustering approach for exploring the feasible/efficient solutions of MCDM problems , 1980, Comput. Oper. Res..
[4] Application of the Monte Carlo method to systems of nonlinear algebraic equations , 1964 .