Differential Evolution: A Practical Approach to Global Optimization (Natural Computing Series)
Differential Evolution: A Practical Approach to Global Optimization (Natural Computing Series)
Evolutionary programming made faster
IEEE Transactions on Evolutionary Computation
A robust stochastic genetic algorithm (StGA) for global numerical optimization
IEEE Transactions on Evolutionary Computation
Neural Network Learning With Global Heuristic Search
IEEE Transactions on Neural Networks
| Hi-index | 0.00 |
Several metaheuristic approaches for global optimization (GO) are investigated and their performances compared in this paper. We critically review and analyze recently proposed Stochastic Genetic Algorithm (StGA) for GO and compare it with our GO hybrid metaheuristic called Genetic LP茂戮驴and Simplex Search(GLP茂戮驴S), which combines the effectiveness of Genetic Algorithms during the early stages of the search with the advantages of Low-Discrepancy sequences and the local improvement abilities of Simplex search. For comparison purposes we also use Fast Evolutionary Programming (FEP) and Differential Evolution (DE) methods. In parallel to our method, FEP and DE, we investigate further, re-run and test the StGA implementation on a number of multimodal mathematical functions. The obtained StGA results demonstrate inferior performance (compared with our GLP茂戮驴Sand DE methods), producing much worse than the reported in [1] results (with the only exception for the two-dimensional cases). We argue that the published in [1] accuracy and convergence speed results (given as number of function evaluations) are incorrect for most of the testing functions and investigate why the method is failing in those cases.