An evaluation of musical score characteristics for automatic classification of composers

An approach that classifies composers of classical music based on certain low-level characteristics of their compositions is proposed. The proposed composition characteristics are descriptive features derived from the time-ordered sequence of pitches in a composition. The notation of a musical score needs to be converted to different syntaxes that are suitable for machine-learning classifiers. There are twelve pitch class features, each specifying the number of occurrences of one pitch class in the score, divided by the total number of notes in the score. CHECKUP, a unique classifier tool that can learn from both vector and regular features and construct mathematical vector relations was employed. The behavior of the feature contribution to accuracy depends on the instrumentation such as all features in Haydn-Mozart with the non-instrument- specific data set contribute to accuracy, except the note duration features, which reduce the accuracy.