Phillip B. Kirlin

Analysis of analysis: Using machine learning to evaluate the importance of music parameters for Schenkerian analysis

While criteria for Schenkerian analysis have been much discussed, such discussions have generally not been informed by data. Kirlin [Kirlin, Phillip B., 2014 “A Probabilistic Model of Hierarchical Music Analysis.” Ph.D. thesis, University of Massachusetts Amherst] has begun to fill this vacuum with a corpus of textbook Schenkerian analyses encoded using data structures suggested byYust [Yust, Jason, 2006 “Formal Models of Prolongation.” Ph.D. thesis, University of Washington] and a machine learning algorithm based on this dataset that can produce analyses with a reasonable degree of accuracy. In this work, we examine what musical features (scale degree, harmony, metrical weight) are most significant in the performance of Kirlins algorithm.

Probabilistic Generation of Ragtime Music from Classical Melodies

This paper examines the computational problem of taking a classical music composition and algorithmically recomposing it in a ragtime style. Because ragtime music is distinguished from other musical genres by its distinctive syncopated rhythms, our work is based on extracting the frequencies of rhythmic patterns from a large collection of ragtime compositions. We use these frequencies in two different algorithms that alter the melodic content of classical music compositions to fit the ragtime rhythmic patterns, and then combine the modified melodies with traditional ragtime bass parts, producing new compositions which melodically and harmonically resemble the original music. We evaluate these algorithms by examining the quality of the ragtime music produced for eight excerpts of classical music alongside the output of a third algorithm run on the same excerpts; results are derived from a survey of 163 people who rated the quality of the ragtime output of the three algorithms.