The selection of musical features plays an important role in a variety of automated music analytic tasks such as music transcription, recommendation, classification, playlist generation, performance evaluation and music information retrieval. Music data available in various digital forms, such as metadata, notations, audio files, online comments and tags and lyrics, is used as an input to the music analyser system. Various approaches are used on music digital data to extract features for identifying relevant information. Feature selection is a critical step in the success of a music analysis task, which is automated or semi-automated using a different machine learning approach. To understand the significance of feature selection, automatic music emotion recognition using supervised machine learning is explored as a case study. Challenges in automating such tasks in computational musicology are discussed so as to give the reader an overview of the complexity in this interdisciplinary research domain.
CITATION STYLE
Velankar, M., & Kulkarni, P. (2023). Music Feature Extraction for Machine Learning (pp. 59–70). https://doi.org/10.1007/978-981-99-0887-5_4
Mendeley helps you to discover research relevant for your work.