Lower bounds for randomized direct search with isotropic sampling |
| |
Authors: | Jens Jägersküpper |
| |
Institution: | Technische Universität Dortmund, Informatik 2, 44221 Dortmund, Germany |
| |
Abstract: | Randomized direct-search methods for the optimization of a function f:Rn→R that is given by a black box for f-evaluations are investigated. These iterative methods generate new candidate solutions by adding isotropically distributed vectors to the current candidate solution. Lower bounds on the number of f-evaluations necessary for reducing the approximation error in the search space are proved. |
| |
Keywords: | Heuristic optimization Direct search Random search Probabilistic analysis Theory |
本文献已被 ScienceDirect 等数据库收录! |