Automatic Tonal Music Composition Using Functional Harmony

摘要

The application of Artificial Intelligence technology to the field of music composition has always been fascinating. Different algorithms are created for automatic music composition and in all cases it could be possible that the sequence of the notes of the melody doesn't permit to obtain a correct sequence of the chords (building on the base of the notes of the melody) on the base of the musical grammar. This paper, which outlines key ideas of our research in this field, provides a step to pass this gap: it proposes a method based on a self-learning model that combines De La Motte's theory of Functional Harmony in a Markov process.