Abstract
Music is a tool that has been integrated into society for thousands of years; it has influenced social aspects of life and has also aided in communication. Today we have various uses for music that go past our traditional uses for entertainment and self-expression. For example, music therapy has been seen to show improvements in patients with Alzheimer’s disease, depression, and PTSD. Additionally, music has played a role in political movements, demonstrating its emotional power. Social media relies heavily on the music industry as many social media posts include music either in the background, or as the forefront of posts. With this broad spectrum of usage, it is important that music is easily organized and accessible. Music genre classification driven by machine learning is a way to keep up with the rapid growth of available music. This technology can be used for the recommendation systems that music streaming services use as well as furthering the medical uses of music. Current classification models struggle with reliability, diversity of datasets, and applicability to real world usage. Future work includes developing diversified datasets and allowing models to classify music genres from all over the world as well as developing increasingly genre-specific algorithms to improve overall performance. The goal of this thesis was to create genre classification models that show promising outcomes to assist the growth of the field. Three machine learning models were created: a K-nearest neighbor model, a convolutional neural network, and a recurrent neural network. All of the models achieved high accuracy on par with previous published works. The K-nearest neighbor model achieved the best performance with an accuracy of 96.5%. As the way we interact with music continues to evolve, it is important to develop adequate tools to further technological advances that expand the role of music in society.
Advisor
Rushit Dave
Committee Member
Rebecca Bates
Committee Member
Sun Kyeong Yu
Committee Member
Mansi Bhavsar
Date of Degree
2025
Language
english
Document Type
Thesis
Degree Program/Certificate
Cognitive Science, with an emphasis in Computer Science
Degree
Bachelor of Science (BS)
Department
Computer Information Science
College
Science, Engineering and Technology
Recommended Citation
Kidanue, R. (2025). Machine learning-driven music genre recognition [Bachelor of Science thesis, Minnesota State University, Mankato]. Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato. https://cornerstone.lib.mnsu.edu/undergrad-theses-capstones-all/8/