Musiplexity
My undergraduate thesis research project, which uses Python libraries to test if computers are able to sort music by genre.
Topological Data analysis & Persistent homology
A field in mathematics that encompasses different disciplines, such as data analysis and machine learning algorithms.
A method within topological data analysis called persistent homology finds features in structures within graphical planes. The holes and structures that are found within these planes will give them a unique identity and allow researchers to categorize large datasets into any classification that is required.
Vietoris-Rips Complex
A complex within persistent homology that explores holes and structures within datasets.
The complex is used in a way to examine holes and structures that are created at different radii. The question is, What kind of meaning do these holes and structures carry as the radius increases? That is exactly what the Vietoris-Rips show us when we use this method.
A machine learning algorithm that finds the centroids within a cluster of data.
Once the song has been broken up into datasets, K-means clustering is used to determine whether the datasets resemble one genre over another.
Thesis report
If you would like to learn more about the thesis, click above. This report includes snipits of code demonstrating the use of Python as well as further explanations of the math behind the code.