Thesis

For my thesis, I contributed two new methods to the tidyclust package in R. The tidyclust package exists as a part of the tidyverse collection of packages to unify implementations for unsupervised learning (clustering) methods under a common interface that is user-friendly and follows design principles from the tidyverse. The two methods I added were density-based clustering using the DBSCAN algorithm and model-based clustering using Gaussian mixture models. For this work, I had to research each respective method and write code to bring each method into tidyclust, justifying my design choices along the way, which you can read more about in my paper or see in my presentation!