Predicting the Composer and Style of Jazz Chord Progressions

Abstract

Jazz music is a genre that consists mainly of improvising over known tunes, represented as a lead sheet. This study addresses the question ‘to what extent does a lead sheet carry information about its composer?’ Primarily, this study considers chord progressions alone, and secondarily melodic and temporal information combined with various multiple viewpoint models. Using these classifiers, a novel subsequence selection algorithm is presented to trace stylistic similarities within a lead sheet. We conclude that composers can, to a reasonable extent, be recognized from their chord progressions, and that the consideration of melodic and temporal information improves classification accuracy by a small but statistically significant amount.

Keywords:

• harmony
• Markov models
• prediction
• multiple viewpoints
• jazz
• classification

Acknowledgements

The authors would like to thank Daniel Martín, Jeff Suzda and Marcus Pearce for their contributions to the study.

Notes

1 where the average observed classification accuracy , standard deviation , is obtained over repeated runs and compared to , the null hypothesis equating to the baseline accuracy.

2 All corrected -values are Bonferroni corrected by dividing the significance level, , by the number of simultaneous hypotheses.

This research was conducted within the Flow Machines project which received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013)/ERC Grant Agreement no. 291156.