Harmonising Chorales by Probabilistic Inference

We describe how we used a data set of chorale harmonisations composed by Johann Sebastian Bach to train Hidden Markov Models. Using a probabilistic framework allows us to create a harmonisation system which learns from examples, and which can compose new harmonisations. We make a quantitative comparison of our system's harmonisation performance against simpler models, and provide example harmonisations.

• Type: Conference
• Year: 2004
• Where: NIPS
• Authors: Moray Allan, Christopher K. I. Williams
• Keywords: Hidden Markov Models, Johann Sebastian Bach, NIPS 2004, NIPS 2007, System's Harmonisation Performance