wolny / Fitness-Deterioration

Population clustering and objective function deterioration

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Configuration file: application-context.xml
Build project: mvn clean install
Program execution: mvn exec:java

Program performs global optimization for a given problem.
Global optimization algorithm is controled by a hybrid approach
called 'Sequantial Niching'. It performs several runs of specified
EA. After each run the algorithm alters fitness function in a way
that prevents exploration of optimas founded in previous runs.

Program writes the results in 'diagrams' catalog:
* cluster's members are stored in files of the name 'clusterN', where N is the cluster's number
* populations for each run are stored in files 'populationN', where N is the number of EA run
* fitness landscape before the algorithm start is stored in file 'originalFitnessLand'
* reachablity plot for OPTICS algorithm is stored in 'reachabilityPlot'
* fitness landscape after each run is stored in 'fintessLand_iterN' where N is the number of EA run


GNUPLOT tips:
set pointsize 0.5
set pm3d map

# plot cluster files whith 'pointtype 7'
# plot fitness landscape with 'with pm3d'
splot 'file' wiht pm3d

#plotting pm3d and population on the same plot
splot 'fitnessLand_iter1' with pm3d, 'fakePopulation1' pointtype 7 notitle

  

About

Population clustering and objective function deterioration

License:MIT License


Languages

Language:Java 99.4%Language:Shell 0.6%