My first academic workshop, Encore 2025
My senior capstone for my Data Science bachelor's degree is focused on Differential Privacy. We get to meet with our mentor, Yu-Xiang Wang, once a week to discuss our progress across two quarters. My group got lucky because he was teaching a graduate course on Differential Privacy the same time as our capstone, which meant that we got to learn a lot more about the topic than the typical data science student.
The luck didn't stop there. At the start of this quarter (Winter 2025), we were invited to attend the Workshop on Defining Holistic Private Data Science for Practice which just so happened to be hosted here at UCSD by Encore. This was an excellent opportunity to dive even deeper into Differential Privacy and be surrounded by those who wrote the papers that taught me Differential Privacy.
Setting the scene
The workshop was a three day event specifically on getting Differential Privacy into practice. This wasn't a realm for undergrads. I was surrounded by professors and industry professionals. Harvard, Berkeley, Google, you name it. The workshop was somewhat intimidating at first, but I quickly found that everyone was friendly and welcoming of my questions. There was certainly a bit of shock factor when I initially introduced myself as an undergrad.
Filling in the gaps
As I watched the talks, the role of Differential Privacy in the world became clearer and clearer. My biggest realization was about just how shaky the legs of Differential Privacy were. The math is sound but getting the theory into practice was a much rougher road. Questions about whether we needed new definitions for necessary conditions were asked and solutions were proposed. The promise that epsilon guarantees suddenly became problematic. Epsilon is unintuitive and hard to understand. How do we explain the guarantees to lawmakers and regulators? Can we use a different metric? Do we need to explain it better? All new questions to me and it was exciting to be in the room with the people who were asking and would be answering these questions.
My reflection of the workshop
Being in such an academic environment was a great experience. So incredibly different was the atmosphere of the workshop compared to that of my classes and degree as a whole. I feel incredibly fortunate to have been able to attend this workshop and I'm eager to keep a keen eye on the future of Differential Privacy.
The data science capstone thus far has been an extremely beneficial experience. Our capstone showcase is coming up on Pi Day, March 14th 2025. I look forward to contributing a small part to the field of Differential Privacy.