Performance of a Bell-Curve Based Evolutionary Optimization Algorithm

@inproceedings{SobieszczanskiSobieski2000PerformanceOA,
  title={Performance of a Bell-Curve Based Evolutionary Optimization Algorithm},
  author={Sobieszczanski-Sobieski and klin. Ass. Dr. Roman Laba and Rex K. Kincaid},
  year={2000}
}
we examined the eeect of parameter changes on the performance of BCB. Three rules of thumb were developed. First, we found that it was crucial to scale the values of the decision variables. Second, we found that, in general, r >> m is best. This is partially due to the innuence of the Euclidean distance between parents on the placement of the center of the n ? 1 dimensional hypersphere in addition to m. Third, we found that the performance of BCB is more sensitive to the value of penalty{2 than… CONTINUE READING