Analysis of the Metrics and Evaluations Methods for Music

Ziteng Li

2024

Abstract

As a matter of fact, in recent decades, computer music has grown into a dominant force, revolutionizing both music creation and evaluation methods. This study explores the evolution of music evaluation from traditional, subjective approaches to more systematic, quantitative methods enabled by computational advancements. The research focuses on key evaluation metrics, including emotion, rhythm, and similarity, and how models like N-gram and Hidden Markov Models (HMM) capture melodic and rhythmic features. This research highlights recent progress in using deep learning algorithms for music assessment and their application in tasks like emotion recognition and music recommendation. Despite the successes, existing models often struggle with complex emotional expressions and cross-cultural diversity in music. The findings suggest that future improvements in music evaluation can be achieved through integrating advanced machine learning techniques and multi-modal analysis. These results contribute to the development of more objective and comprehensive evaluation methods, ultimately benefiting various applications in music classification, recommendation, and automated composition.

Paper Citation

in Harvard Style

Li Z. (2024). Analysis of the Metrics and Evaluations Methods for Music. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 198-202. DOI: 10.5220/0013512500004619

in Bibtex Style

@conference{daml24,
author={Ziteng Li},
title={Analysis of the Metrics and Evaluations Methods for Music},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={198-202},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013512500004619},
isbn={978-989-758-754-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Analysis of the Metrics and Evaluations Methods for Music
SN - 978-989-758-754-2
AU - Li Z.
PY - 2024
SP - 198
EP - 202
DO - 10.5220/0013512500004619
PB - SciTePress