On the Use of Matrix Based Representation to Deal with Automatic Composer Recognition
In this article the use of a matrix based representation of pieces is tested for the classification of musical pieces of some well known classical composers. The pieces in two corpora have been represented in two ways: matrices of interval pair probabilities and a set of 12 global features which had previously been used in a similar task. The classification accuracies of both representations have been computed using several supervised classification algorithms. A class binarization technique has also been applied to study how the accuracies change with this kind of methods. Promising results have been obtained which show that both the matrix representation and the class binarization techniques are suitable to be used in the automatic composer recognition problem.
KeywordsMatrices Pairwise classification Composer recognition
This work has been partially supported by the Basque Government Research Teams grant (IT900-16) and the Spanish Ministry of Economy and Competitiveness. TIN2015-64395-R (MINECO/FEDER).
- 7.Goienetxea, I., Neubarth, K., Conklin, D.: Melody classification with pattern covering. In: 9th International Workshop on Music and Machine Learning (MML 2016), Riva del Garda, Italy (2016)Google Scholar
- 10.Hillewaere, R., Manderick, B., Conklin, D.: Global feature versus event models for folk song classification. In: Proceedings of the 10th International Society for Music Information Retrieval Conference, Kobe, Japan, pp. 729–733 (2009)Google Scholar
- 11.Hillewaere, R., Manderick, B., Conklin, D.: String methods for folk tune genre classification. In: Proceedings of the 13th International Society for Music Information Retrieval Conference, Porto, Portugal (2012)Google Scholar
- 12.van Kranenburg, P., Conklin, D.: A pattern mining approach to study a collection of Dutch folk-songs. In: Proceedings of the 5th International Workshop on Folk Music Analysis (FMA 2016), Dublin, pp. 71–73 (2016)Google Scholar
- 13.Mckay, C., Fujinaga, I.: jsymbolic: a feature extractor for midi files. In: Proceedings of the International Computer Music Conference, pp. 302–305 (2006)Google Scholar
- 14.Sapp, C.S.: Online database of scores in the humdrum file format. In: ISMIR 2005, Proceedings of 6th International Conference on Music Information Retrieval, 11–15 September 2005, London, UK, pp. 664–665 (2005)Google Scholar