Agglomerative Clustering

This research was done as part of the Undergraduate Research Opportunities in Computing (UROC) at Indiana University with Tauffique Hussain, as PhD student under Ariful Azad.

The goal of this project was to begin an initial investigation into the development of an iterative clustering method to be used for large, frequently updated graphs. As a first step, we applied an agglomerative clustering method based on Earth Mover’s distance to discrete datasets. We then evaluated the accuracy of clustering at various thresholds.