Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses
Cluster analysis provides insight into musical patterns in composition, performance, and perception. Despite its wide adoption in music research, understanding how specific features affect clustering solutions remains challenging. For example, features such as mode (i.e., major/minor), timing, signa...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SAGE Publishing
2023-12-01
|
| Series: | Music & Science |
| Online Access: | https://doi.org/10.1177/20592043231216257 |
| _version_ | 1797374447651389440 |
|---|---|
| author | Cameron J. Anderson Michael Schutz |
| author_facet | Cameron J. Anderson Michael Schutz |
| author_sort | Cameron J. Anderson |
| collection | DOAJ |
| description | Cluster analysis provides insight into musical patterns in composition, performance, and perception. Despite its wide adoption in music research, understanding how specific features affect clustering solutions remains challenging. For example, features such as mode (i.e., major/minor), timing, signal amplitude, and pitch are often intercorrelated, making it difficult to understand their specific role within different clusters. To demonstrate how accumulated local effects (ALEs) can help with this challenge, here we analyze 48 excerpts from complete sets of preludes by Bach and Chopin, showing how specific features contribute to two- and three-cluster analyses. These exploratory analyses reveal that ALEs can identify salient or subtle data patterns from cluster analyses by tracking how changes in features affect cluster membership. We explore these insights in visualizations quantifying feature importance and an interactive companion application ( https://maplelab.net/feature-importance/ ) featuring the analyzed audio. Following a demonstration of this method, we suggest how it can be applied to explore topics of interest to researchers in music information retrieval, empirical musicology, and music cognition alike. |
| first_indexed | 2024-03-08T19:05:59Z |
| format | Article |
| id | doaj.art-a26b3ab36fee4e61adad69965d9c081c |
| institution | Directory Open Access Journal |
| issn | 2059-2043 |
| language | English |
| last_indexed | 2024-03-08T19:05:59Z |
| publishDate | 2023-12-01 |
| publisher | SAGE Publishing |
| record_format | Article |
| series | Music & Science |
| spelling | doaj.art-a26b3ab36fee4e61adad69965d9c081c2023-12-28T00:03:20ZengSAGE PublishingMusic & Science2059-20432023-12-01610.1177/20592043231216257Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster AnalysesCameron J. Anderson0Michael Schutz1 Department of Psychology, Neuroscience and Behaviour, , Hamilton, Canada School of the Arts, , CanadaCluster analysis provides insight into musical patterns in composition, performance, and perception. Despite its wide adoption in music research, understanding how specific features affect clustering solutions remains challenging. For example, features such as mode (i.e., major/minor), timing, signal amplitude, and pitch are often intercorrelated, making it difficult to understand their specific role within different clusters. To demonstrate how accumulated local effects (ALEs) can help with this challenge, here we analyze 48 excerpts from complete sets of preludes by Bach and Chopin, showing how specific features contribute to two- and three-cluster analyses. These exploratory analyses reveal that ALEs can identify salient or subtle data patterns from cluster analyses by tracking how changes in features affect cluster membership. We explore these insights in visualizations quantifying feature importance and an interactive companion application ( https://maplelab.net/feature-importance/ ) featuring the analyzed audio. Following a demonstration of this method, we suggest how it can be applied to explore topics of interest to researchers in music information retrieval, empirical musicology, and music cognition alike.https://doi.org/10.1177/20592043231216257 |
| spellingShingle | Cameron J. Anderson Michael Schutz Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses Music & Science |
| title | Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses |
| title_full | Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses |
| title_fullStr | Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses |
| title_full_unstemmed | Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses |
| title_short | Understanding Feature Importance in Musical Works: Unpacking Predictive Contributions to Cluster Analyses |
| title_sort | understanding feature importance in musical works unpacking predictive contributions to cluster analyses |
| url | https://doi.org/10.1177/20592043231216257 |
| work_keys_str_mv | AT cameronjanderson understandingfeatureimportanceinmusicalworksunpackingpredictivecontributionstoclusteranalyses AT michaelschutz understandingfeatureimportanceinmusicalworksunpackingpredictivecontributionstoclusteranalyses |