Skip to content

Efficient optimization for any binary domain problem where relationships between variables are known.

License

Notifications You must be signed in to change notification settings

brianwgoldman/GrayBoxOptimization

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

46 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Gray-Box Optimization using the Parameter-less Population Pyramid

To compile you will need C++11. We use gcc version 4.8.2 for our complilation.

Our build system uses Makefiles to build. You can compile the release version by changing directory to Release and calling "make GBO".

All of the source code is available in the 'src' directory.

To run an experiment, call the executable with command line arguments for configuration. This will run the default test configuration:

Release/GBO config/default.cfg

The command line accepts any number of configuration files and configuration flags given in the form "-key value" where key is the name of the option to change and value is the value to change it to. Arguments override in last given order. For example:

Release/GBO config/default.cfg -problem NearestNeighborNKQ -length 1000 -verbosity 2

This line will use the default configuration, setting the problem to Nearest Neighbor NKQ with a genome size of 1000, and increasing the amount of information output to the screen.

Also, see main.cpp for implementation related details.

The data used in the paper is available as part of the GBO-using-P3.tar.gz file. All of the graphing and statistical processing scripts are contained in data_parsing/.