annGO

An existing Java application handling and visualizing Go games between human and computer players (including trained and evolved ANNs) should be improved and extended with a generic input pattern generator for supervised learning of neural Go players. The patterns should be extracted from arbitrary Simple Go Format (SGF) files containing Go games played by amateurs, computers, or professional players. A plug-in architecture should enable experiments with sophisticated board representations in future work. Moreover, ANN training using the generated patterns, and tournaments held between arbitrary computer players are to be implemented.