Can we use machine learning to analyze the DNA of a melody?
In this lecture from MIT's Computational Music Theory and Analysis, guest speaker Kiki Gutierres demonstrates how computational research can be applied to popular music corpora, including the global phenomenon Baby Shark.
The session explores the technical processes of feature extraction and how these methods allow researchers to mathematically analyze similarity and difference within melodic examples. By applying machine learning principles to musical data, the lecture demonstrates how complex auditory patterns can be broken down into quantifiable features for computational study.
Presented as part of the Spring 2023 MIT 21M.383 course, the lecture moves from specific case studies of song corpora to the broader mechanics of analyzing musical structure through a computational lens.
Source: Class 28 Video: Feature Extraction and Machine Learning (II)
