NEUROGEN, musical composition using genetic algorithms and cooperating neural networks
Paul M. Gibson, Jason Byrne · International Conference on Artificial Neural Networks · 1991
NEUROGEN has been designed with a view to producing small diatonic, western-type, four-part harmony compositions using the knowledge extracted from a set of example musical fragments provided by the user. The aim has been to produce a piece of coherent music that resembles that typically found in traditional hymns. A set of neural networks are used to capture the conceptual ideas that build 'good' music and this knowledge is then used to direct a search for the ultimate composition. Genetic algorithms hold many member states representing partial musical fragments. The neural networks cooperate to produce a heuristic value that represents the worth of each of these musical fragments. This value is then used to evolve better compositions based on fragments with high fitness values. The use of Neural Networks as an evaluation function has proved successful in the guidance of genetic algorithms. This success is evident in the type of composition produced by NEUROGEN. >