Generating Music With Graphical Models

Abstract

Music can be seen as a set of pieces of sound, a set of instrument combinations or a set of notes. Modeling a song in one of these ways provides the possibility to study what we are going to play with respect to the previous times. Because of this discrete splitting
of a piece of music, we thought of using graphical models. This thesis is dealing with the use of Markov Models and family in order to model music. We were interested in building a model complex enough to catch the local information of a measure and the more global information of a set of measures.

Date of Award	Sept 2000
Original language	English
Awarding Institution	Aston University