Restricted-Access

 

 
       

 

MMGAs> Fitness sharing: Fitness sharing (FS)

                                                                                                                                                                                                                                                                

 

Last update: 06/2011

 

 

     Parameters

 

 

■ Niche radius (σshare)

Slope of the sharing function (fi) whose most used value is 1

 

 

     Description

 

Classic Fitness Sharing (FS) was the first method on multi-modality in GAs. It is based on the penalization of the areas of the searching space with more solutions within the population. This way, solutions belonging to densely populated areas will see their quality decreased in order to allow the exploration of other sparsely populated areas. The selection process uses the shared fitness of individual i (fi*) defined as the original fitness (fi) divided by the sharing function, Sh(d(i,j)) (see Equation 1):

where the sharing function depends on the distance between individuals i and j (d(i,j)) according to Equation 2.

This method is applied prior to the selection process. Firstly, it starts with the measurement of the distance between the first solution in regard to the rest of the solutions of the population, calculating the accumulated value of its sharing function. Once the test of all the population is finished the shared fitness is calculated by dividing the original quality by the sharing function. This process is repeated with all the solutions, until every modified quality is obtained, and at this point the parents selection is performed.

 

This process is shown in the following flowchart:

 

 

The main disadvantage of the methods based on the niche radius is the a priori ignorance of the most suitable value of this parameter. Therefore a prior experimentation is necessary to adapt the method to each type and size of problem.

 

Later studies proved that this method had some restrictions. Firstly, it should be applied with the less biased selection methods, as for instance with the stochastic universal selection (SUS) or with the stochastic remainder selection (SRS). Secondly, this method is not able to maintain the detected niches in the first generations stable during later generations in problems with high difficulty (whether by a large number of optima or by deception in the landscape).

 

 
 

     Executable

 

The following three files have to be in the same folder. Do not change the name of the file. Double-click the .exe file to run it.

 

  JSSP

Executable

exe

Data File

Open .txt file of the instance you want to run (for JSSP), copy and paste its contents inside this file.

D_Entrada

Parameters File

Open the file and change the parameters

D_Algoritmo

 

The following five files of solutions are generated:

D_Sal_Mejores-txt (includes the best solutions).

D_Sal_Optimos (includes the value of the objective function, i.e., Cmax).

D_Sal_NumOptimos (includes the number of different solutions where the optima value specified in the data file has been reached). This file is empty if this value is not reached.

D_Sal_Distancias (includes the distances between solutions where the optima value specified in the data file has been reached). This file is empty if either this value is not reached or only one solution is achieved.

D_Sal_Soluciones (includes the schedule of the solutions where the optima value specified in the data file has been reached). This file is empty if this value is not reached.

 

 

     References

 

Goldberg, D. E. and Richardson, J. (1987). Genetic algorithms with sharing for multimodal function optimization. In Proc. of the 2nd International Conference on Genetic Algorithms,  41-49.

 

 

   

Universidad de Valladolid. Webmaster Marta Posada