This is a project I co-created with Austin Taylor at the end of a scientific computation course. Inspired by the desire to reduce the weekly workload in clustering drivers to riders in a music performance club I was in where we would do weekly performances for local homeless shelters and retirement centers.
In essence, we coded an agglomerative clustering method (complete-linkage clustering) from scratch because it fit our use case quite well to make nicely compact rider groups. The information was collected through google forms and downloaded as a CSV from a weekly email sent out. (Fake data provided to supplement the proof-of-concept work).
rideshare_clustering.pdf - compiled from Latex, this pdf was the original project submission (spelling errors and all)
You can read more about the project on my blog