Maurice Williams graduated from the School of Information with a Master of Information and Data Science in 2020. He currently works as a Predoctoral Research Fellow at the Stanford Institute for Economic Policy Research. Prior to the I School, Maurice graduated with a bachelor’s degree in economics from UC Berkeley.
Why did you choose the 5th Year MIDS program?
I liked the idea of completing my master’s in just a year, and the MIDS for Life promise of providing access to all current and future course content made it particularly attractive.
What made the I School special?
I’m an economics nerd, so knowing that Google's Chief Economist Hal Varian was the I School’s first dean was a big selling point for me. It told me that the I School wasn’t just a place for software engineers, but was for a wider, more interdisciplinary approach using the topics and tools I was interested in.
What was your favorite class and why?
I really enjoyed W266: Natural Language Processing and Deep Learning with Matthew Sims. It was my first time diving into deep learning architectures, and it was highly stimulating. What made it really great, though, was the final project. My project partner and I were able to do something less traditional in this space, trying to leverage these tools for inference and insight, as opposed to creating an application of sorts. That kind of freedom isn’t something unique to only this course — instructors encourage you to take chances with most final projects at MIDS.
Any plans for the future?
I’m currently a Predoctoral Research Fellow at the Stanford Institute for Economic Policy Research, which is my last stop on the way to a Ph.D. program in Economics. The team I’m on has a workflow exactly like a data science team does. We use git and GitHub for all our projects and almost everything is in Python and R; we leverage cloud resources for parallel processing big jobs, and we’re expected to leave clean, commented and well-documented code. I’m well prepared for the day-to-day work as a researcher because of MIDS.
