1. Introduction - richoux/GHOST GitHub Wiki

GHOST (General meta-Heuristic Optimization Solving Tool) is a user-friendly C++ framework designed to help C++ developers to model their own optimization problems.

This is a lightweight, dependence-free library aiming for speed. GHOST contains a very efficient local search solver with only few parameters (for which default values usually lead to state-of-the-art results), evicting the tedious parameter-tuning process usually mandatory with such kind of solvers. It also contains a non-backtracking complete solver that can be used to find all solutions of a given problem, if needed.

Initially, GHOST has been developed to handle optimization problems within StarCraft: Brood war. It is a generalization of the Wall-in project (see github.com/richoux/Wall-in). Now, GHOST is a general framework aiming to model and solve any kind of combinatorial optimization problems.

Source code and documentation

The source code is available at github.com/richoux/GHOST under the terms of the GNU GPL v3 license, and the full doxygen documentation is available at richoux.github.io/GHOST.

Possibilities for dual-license

GHOST is under the terms of the GNU GPL v3 license. However, I can propose a dual-license: I already sold a proprietary version of GHOST to the game studio Insane Unity for their game Win That War!, via a technological transfer. Feel free to contact me if you are interested by GHOST under a proprietary license: [email protected]

Scientific papers about GHOST:

• Florian Richoux, Alberto Uriarte and Jean-François Baffier, GHOST: A Combinatorial Optimization Framework for Real-Time Problems, IEEE Transactions on Computational Intelligence and AI in Games, 2016.
• Julien Fradin and Florian Richoux. Robustness and Flexibility of GHOST, AIIDE Third Workshop on Artificial Intelligence in Adversarial Real-Time Games, 2015.

Related scientific papers:

• Florian Richoux, Alberto Uriarte and Santiago Ontañón, Walling in Strategy Games via Constraint Optimization, 10th Annual AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE 2014), 2014.