A hybrid approach to supervised machine learning for algorithmic melody composition

TL;DR

This paper introduces a hybrid machine learning algorithm combining higher-order Markov models and melody contour analysis to generate monophonic melodies that mimic a given style, validated through online listening tests.

Contribution

It presents a novel hybrid approach integrating parametric Markov models and contour concepts for improved melody composition.

Findings

Enhanced melody quality with context-aware Markov models

Significant improvement shown in online listening tests

Method effectively captures stylistic features of sample melodies

Abstract

In this work we present an algorithm for composing monophonic melodies similar in style to those of a given, phrase annotated, sample of melodies. For implementation, a hybrid approach incorporating parametric Markov models of higher order and a contour concept of phrases is used. This work is based on the master thesis of Thayabaran Kathiresan (2015). An online listening test conducted shows that enhancing a pure Markov model with musically relevant context, like count and planed melody contour, improves the result significantly.