Thesis

For my thesis, I contributed new functionality to the tidyclust package in R. The tidyclust package exists as a part of the tidymodels framework , which follows the principles of the tidyverse to establish a consistent and reproducible workflow for unsupervised learning algorithms. My work adapted the Apriori and ECLAT algorithms for frequent itemset mining. This involved establishing a cluster assignment strategy that groups items based on frequent itemsets and support values, as well as a standardized methodology for predicting missing items. You can read more about my design choices in my paper or see them in my presentation!