Download C++ implementation Download MATLAB implementation Published applications References under construction Stochastic optimization tool CERAF strategy

• multistart strategy with memory enabling escape from local extremes
• niching strategy
• suitable complement for evolutionary algorithms
• To conclude the algorithm description, let's sumarize algorithm's control parameters:
  
  

  parameter	description	recommended value
  RAD	control the radius of the radioactivity area	0.25
  deact_rate	control the decreasing of size of radioactive area	0.995
  quiet	control number of generation before new local extreme is marked	100

  
  

Published applications

• Implementation with SADE algorithm was tested on a set of twenty functions (ANDRE) and compared with Standard Binary Genetic Algorithm (SBGA), Extended Binary Genetic Algorithm (EBGA), Differential Evolution (DE) is presented in [1]
• Implementation with GRADE algorithm was tested on a set of twenty functions (ANDRE) and compared with SADE algorithm, GRADE algorithm alone and with GRBFN strategy in my Ph.D. thesis [2]
• Implementation with GRADE algorithm was applied to neural network training and presented in [3]

C++ implementation

• it is considered for maximization of functions
• C++ implementation of CERAF strategy consists of 2 following files:
ceraf.h(826B, version 0.0.1, released 16 Mar 2009) - definition of CERAF class
ceraf.cpp(2.11kB, version 0.0.1, released 16 Mar 2009) - definition of CERAF class functions
• C++ version of CERAF algorithm was implemented to GRADE algorithm, all necessary files of CERAF strategy together with GRADE algorithm, files specifying cost function to be minimized and example of execution could be downloaded as a following archive: grade+ceraf-cpp.zip (8.09kB, released 16 Mar 2009)

Matlab implementation

• this implementation is developped for Matlab 7.6.0 and it uses randsample function from Statistics Toolbox
• it is considered for minimization of functions
• matlab implementation of CERAF strategy consists of 2 following files:
CERAFinit.m(1.46kB, version 0.0.5, released 24 Mar 2009) - initialization script of the strategy
CERAFcheck.m(2.25kB, version 0.0.6, released 6 Sep 2015) - script checking, if new local extreme wasn't found and if any new solution is not in zone of existing local extremes
• matlab version of CERAF algorithm was implemented to GRADE algorithm, all necessary files of CERAF strategy together with GRADE algorithm, files specifying cost function to be minimized and example of execution could be downloaded as a following archive: grade+ceraf-matlab.zip (6.47kB, released 24 Mar 2009)

References