PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Comparing probabilistic models for melodic sequences
Athina Spiliopoulou and Amos Storkey
Lecture Notes in Computer Science Volume 6913, pp. 289-304, 2011.

Abstract

Modelling the real world complexity of music is a challenge for machine learning. We address the task of modelling melodic sequences from the same music genre. We perform a comparative analysis of two probabilistic models; a Dirichlet Variable Length Markov Model (Dirichlet-VMM) and a Time Convolutional Restricted Boltzmann Machine (TC-RBM). We show that the TC-RBM learns descriptive music features, such as underlying chords and typical melody transitions and dynamics. We assess the models for future prediction and compare their performance to a VMM, which is the current state of the art in melody generation. We show that both models perform significantly better than the VMM, with the Dirichlet-VMM marginally outperforming the TC-RBM. Finally, we evaluate the short order statistics of the models, using the Kullback-Leibler divergence between test sequences and model samples, and show that our proposed methods match the statistics of the music genre significantly better than the VMM.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Article
Additional Information:Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases, Part III, ECML PKDD 2011, Athens, Greece, September 5-9, 2011
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:8595
Deposited By:Athina Spiliopoulou
Deposited On:14 February 2012