Genetic algorithms in musical style oriented generation
Pauli Laine, Mika Kuuskankare · 2002
Several different approaches have been used to explore style oriented music generation. We describe a system where music generating functions are automatically produced with help of genetic algorithms. Genetic algorithms are search and machine learning procedures that are based on natural selection. Genetic algorithms such as any other algorithms can be used to produce musical output. There are few basic approaches when using genetic algorithms: they can be used as an optimising method to solve complex parameterised and/or rule based problems. Genetic algorithms are also widely used on solving problems by generating functions that will perform the selected task the most effectively. This concept is explored in this contribution.>