de León P.J.P., Iñesta J.M. (2004) Musical Style Classification from Symbolic Data: A Two-Styles Case Study. In: Wiil U.K. (eds) Computer Music Modeling and Retrieval. CMMR 2003. Lecture Notes in Computer Science, vol 2771. Springer, Berlin, Heidelberg
In this paper the classification of monophonic melodies from two different musical styles (Jazz and classical) is studied using different classification methods: Bayesian classifier, a k-NN classifier, and self-organising maps (SOM). From MIDI files, the monophonic melody track is extracted and cut into fragments of equal length. From these sequences, A number of melodic, harmonic, and rhythmic numerical descriptors are computed and analysed in terms of separability in two music classes, obtaining several reduced descriptor sets. Finally, the classification results for each type of classifier for the different descriptor models are compared. This scheme has a number of applications like indexing and selecting musical databases or the evaluation of style-specific automatic composition systems.
music information retrieval self-organising maps bayesian classifier nearest neighbours (k-NN) feature selection