EEBIC Group


GP Music System Page

by Brad Johanson, Supervised by Riccardo Poli



This page contains material related to the GP-Music System, a program which evolves musical sequences.  The system can be used with a human serving as the fitness function, or information on human ratings of musical sequences can be used to train neural networks which rate sequences instead of the human.

The work on the system was done in two parts.  The initial implementation was done during the Spring of 1997 at the University of Birmingham as a mini-project in the MSc Advanced Computer Science program.  Improvements to the system and the development of the neural network automatic fitness raters were done as a Final Project for the Masters program.  Resources associated with the two projects are listed below:

The GP-Music System Mini-Project

The GP-Music System with Automatic Fitness Raters Final Project

  • Sample Code

  • Mantained by Brad Johanson <bjohanso@stanford.edu>