Rosenkoenig

The project Rosenkoenig was created as part of the subjects Neuroinformationsverarbeitung and Genetische Algorithmen at the HTW Dresden.

See:
Description

Packages

de.htwdd.rosenkoenig	de.htwdd.rosenkoenig is the basic package of the Rosenkoenig project.
de.htwdd.rosenkoenig.ga	Contains a Genetic Algorithm to determine an appropriate structure of a Neural Network that could be applied to an artificial player.
de.htwdd.rosenkoenig.game	This packages contains all classes that implement the "business logic" of Rosenkoenig.
de.htwdd.rosenkoenig.gui	All classes that have to do anything with user interaction or game in- and outputs are located here.
de.htwdd.rosenkoenig.gui.genuts	This package contains a GUI implementation based on the Java gui framework Genuts.
de.htwdd.rosenkoenig.neuro	Contains all parts of Rosenkoenig that deal with neural networks.
de.htwdd.rosenkoenig.neuro.net	The package net contains implementations, examples and utilities to create and train neural networks.
de.htwdd.rosenkoenig.neuro.net.examples	This package contains a number of examples to demonstrate the use of the different neural network classes.
de.htwdd.rosenkoenig.neuro.net.training	This package contains classes to train neural networks.
de.htwdd.rosenkoenig.util	Some utility classes are located in this package.

The project Rosenkoenig was created as part of the subjects Neuroinformationsverarbeitung and Genetische Algorithmen at the HTW Dresden.
Rosenkoenig is the Java implementation of the German table top game Rosenkönig, developed by Dirk Henn.

The basic implementation of Rosenkoenig is playable by two human players, but the project also contains an implementation of an artificial player, that takes its decisions based on a Neural Network. To determine an approprite structure for the neural network, a Genetic Algorithm was used.

This project consists of:

• a logical implementation of the Rosenkönig rules
• a GUI based on the Genuts library
• a basic implementation of Feed Forward and Kohonen Neural Networks
• an artificial player based on Neural Networks
• a genetic algorithm to determine the network structure

Rosenkoenig was developed by Thomas Egerer and Dominik Hölz.
The project makes use of GaFrame, a framework for Genetic Algorithms developed by Dominik Hölz and Wolfram Wagner.