An artificial brain to play GO

Program Code: 
1885
Contact: 

Prof Hussein Abbass (h.abbass@adfa.edu.au)

Description of Work: 

Objectives:

Developing a computer system that is capable of playing GO at a professional level is one of the greatest challenges facing artificial intelligence research at the moment. This project will look at advanced artificial intelligence techniques to build a humancompetitive professional GO player. The project will develop a large-scale neuralbased system to model strategies at both the tactical and strategic levels of GO. The research will focus on a computational neural architecture for GO inspired by a number of models in human psychology.

Description of Work:

  • Understanding neural architectures, strategies, and thinking models
  • Understanding a number of architectures and models in Cognitive Science
  • Developing a large-scale neural architecture for GO
  • Developing multi-stage methodologies for training the architecture
  • Testing the competitiveness of the GO player