Education Icon Research Icon Students Icon Community Icon General Support Icon Gift Cart Icon
Skip to main content
A person standing in front of a rack of computers.

Research and Development

With help from PwC, UCSB students tackle a real-world data science problem

The UCSB Data Science Initiative strengthens UC Santa Barbara's visibility as a leader in the field of data science and equips students with tools and techniques that can lead to innovations in business, academia, and society. Thanks to private sector partners such as PricewaterhouseCoopers (PwC), a major international consulting firm, data science students also gained insights into real-world constraints and data.

Over the last academic year, PwC’s team mentored a group of upper-division students working on a Data Science Capstone, an elective three-quarter course that introduces them to key data science concepts through case studies, the application of ethics frameworks and insights into the effective communication of results. The second and third quarters of the course are dedicated to a team project with industry partners and research labs around campus.

"The involvement of sponsors like PwC is integral to the success of the data science capstone program,” said Professor Alex Franks. “The students working with PwC gained valuable experience working with a team to solve problems of real-world relevance."

Robert Bernard is the director of PwC’s Climate Risk Modeling practice, which works with clients to predict weather-related risks from the effects of climate change. He presented the UC Santa Barbara team with a compelling challenge. PwC’s team of data scientists, computer scientists, and software engineers develop very large machine learning models for a variety of industries. If that model could be compressed, it would run faster, consume fewer resources, and be more accessible.

While data compression occurs in all media — from the lossless music files on streaming services to the pictures on mobile phones — model compression is a brand new field. How could one take a supercomputer model that consumes vast disc space and memory and compress it so that it could operate without internet, or on a battery, while maintaining performance?

“This project was great because I started off knowing a little bit more than the students did. By the end, they blew me away. I learned so much from them. It was a great partnership for UC Santa Barbara and PwC,” said Rob.

The students were as excited about the project as they were to work with Rob. Department of Statistics and Applied Probability graduate student Erika McPhillips mentored the undergraduate team.

“Undergraduate and graduate students are kind of still trying to figure out what they want to do,” said Erika. “This project opened my eyes to new industry experiences — so not only undergraduates benefit from this.”

Student team member Yan Lashchev ’23 is a double major in applied mathematics and statistics and data science. He is minoring in linguistics, specifically language and speech tech. Yan’s experience with PwC helped refine his vision for his future.

“I was aware companies had sectors for ‘research & development,’ but I never imagined it would be so similar to the academic research process. This was certainly very exciting for me because I want to continue to learn throughout my career, and this type of position would be a perfect fit. An experience like this gives you a big leg-up compared to other graduates when applying to full-time positions. My long-term goal is to obtain a Ph.D. and that still stands, as I have seen first-hand its usefulness and success in the industry world,” said Yan.

PwC leveraged the students’ research by inviting data scientists across the firm to attend the final presentation on the emerging field of model compression.

“Now we have a white paper and a presentation to refer to. The UC Santa Barbara students relieved the burden of weeks of research from PwC. When we do pursue this, we’ll be ready,” said Rob.

Support UCSB Capstones

PwC has continued to support UC Santa Barbara capstones, now working with the Department of Computer Science on modeling wildfires with machine learning. Learn more about how you can partner with UCSB scholars.