Testing genetic algorithm recombination strategies and the normalized compression distance for computer-generated music

Authors:
Manuel Alfonseca;Manuel Cebrián;Alfonso Ortega
Affiliations:
Escuela Politécnica Superior, Universidad Autónoma de Madrid, Madrid, Spain;Escuela Politécnica Superior, Universidad Autónoma de Madrid, Madrid, Spain;Escuela Politécnica Superior, Universidad Autónoma de Madrid, Madrid, Spain
Venue:
AIKED'06 Proceedings of the 5th WSEAS International Conference on Artificial Intelligence, Knowledge Engineering and Data Bases
Year:
2006

Citing 5
Cited 1

An introduction to Kolmogorov complexity and its applications

An introduction to Kolmogorov complexity and its applications
Automatic composition of music by means of grammatical evolution

APL '02 Proceedings of the 2002 conference on APL: array processing languages: lore, problems, and applications
The similarity metric

SODA '03 Proceedings of the fourteenth annual ACM-SIAM symposium on Discrete algorithms
Algorithmic Clustering of Music

WEDELMUSIC '04 Proceedings of the Web Delivering of Music, Fourth International Conference
Clustering by compression

IEEE Transactions on Information Theory

The Normalized Compression Distance as a Distance Measure in Entity Identification

ICDM '09 Proceedings of the 9th Industrial Conference on Advances in Data Mining. Applications and Theoretical Aspects

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper analyzes the use of the normalized compression distance as a fitness function for the automatic generation of music by means of genetic algorithms, and tests the effect on performance of several genetic recombination procedures. The minimization of the distance of the generated music to a set of musical guides or targets makes it possible to obtain computer-generated music that reminds the style of a certain human author. In spite of the simplicity of the algorithm, the procedure obtains interesting results. The paper includes some considerations on the use of procedures that improve the performance of heavy-tailed distribution processes.