Differential Evolution Estimator

The Differential Evolution estimator is a stochastic global optimization method, a simple and efficient heuristic for global optimization over continuous spaces. More information about this method can be found in the following publication:

  1. Storn, “Differential evolution - a simple and efficient heuristic for global optimization over continuous spaces,” J. Global Optim. 11(4), 341 - 359 (Dec. 1997).

Below is a list of parameters that a user can adjust when using the Differential Evolution method.

DEDiscover uses the method implemented by Adrian Michel and available on github.

Control parameters for the Differential Evolution Estimator
Name Description Constraints and Notes
Max Generations Maximum number of generation before termination Must be >= 0.
Population size Number of candidates created at each generation Must be >= 0.
Mutation Strategy A number between 1 and 5. The description for each strategy will be added when it is located.  
Crossover Factor   Between 0 and 1
Weight   Between 0 and 1