Comparing probabilistic models for melodic sequences

Authors:
Athina Spiliopoulou;Amos Storkey
Affiliations:
School of Informatics, University of Edinburgh, United Kingdom;School of Informatics, University of Edinburgh, United Kingdom
Venue:
ECML PKDD'11 Proceedings of the 2011 European conference on Machine learning and knowledge discovery in databases - Volume Part III
Year:
2011

Citing 8
Cited 0

Training products of experts by minimizing contrastive divergence

Neural Computation
Learning the Long-Term Structure of the Blues

ICANN '02 Proceedings of the International Conference on Artificial Neural Networks
Polyphonic music modeling with random fields

MULTIMEDIA '03 Proceedings of the eleventh ACM international conference on Multimedia
A fast learning algorithm for deep belief nets

Neural Computation
Using Machine-Learning Methods for Musical Style Modeling

Computer
Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations

ICML '09 Proceedings of the 26th Annual International Conference on Machine Learning
Factored conditional restricted Boltzmann Machines for modeling motion style

ICML '09 Proceedings of the 26th Annual International Conference on Machine Learning
A stochastic memoizer for sequence data

ICML '09 Proceedings of the 26th Annual International Conference on Machine Learning

Quantified Score

Hi-index	0.00

Visualization

Abstract

Modelling the real world complexity of music is a challenge for machine learning. We address the task of modeling melodic sequences fromthe 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-VMMmarginally 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.