Author: Silas Garrido
Differential Evolution (DE) algorithms search for solutions to multidimensional optimization problems by simulating ideas based on natural selection. Taking these ideas into account it was developed a specific DE [1] to optimize the values of the parameters of concept drift detection methods. The corresponding code can be downloaded via the link below:
Differential Evolution (insert in /moa/tasks)
[1] Silas G. T. C. Santos, Roberto S. M. Barros, and Paulo M. Gonçalves Jr., “A Differential Evolution based Method for Tuning Concept Drift Detectors in Data Streams” in Information Sciences, 2019.
Silas Garrido - sgtcs@cin.ufpe.br