Statistics in Science, Business, and Politics

Portrait of Wed Piyush
Photo by Jacob Van Blarcom, CLAgency student

Have you ever wondered how online retailers like Amazon recommend such relevant and unique products to you? Ved Piyush, a graduate student in the School of Statistics, can tell you exactly how it’s done. Through his experiences in the School of Statistics, Ved has applied his knowledge and expertise as a data scientist intern at Seagate.

Originally from India, Ved completed his undergraduate degree in statistics at the University of Delhi. “My father is a statistician in the Indian government, and I became interested in his work,” Piyush says. His father’s influence, along with the strong statistics program at the university, motivated Ved to further study statistics and its applications in the US.

After applying to many programs, Ved chose the University of Minnesota. “The program here is well-renowned and has excellent faculty,” he says, laughing about how he didn’t consider the weather in his decision. Now in his final year of graduate school, Ved has completed the required master’s coursework and is working on his final project to present to the graduate committee.

Science with Seagate

After moving to Minnesota, Ved quickly became interested in machine learning and its applications within data science. Fortunately, the school has a close relationship with Seagate, a global technology company that has established a scholarship for statistics students here at the University. They send out an assessment for students to work on and, based on their performance, a few are selected for internships. 

Ved was chosen to be a data scientist intern, working primarily on developing a recommendation system for Seagate’s internal use. His internship began last September; however, he became a full-time intern in January.

In his role, he looks at how to build user profiles based on the Seagate engineer’s usage histories. “For example, if you were to buy a coffee machine on Amazon, it may recommend to you a coffee filter,” Ved says. This process of making recommendations based on an engineer’s likes and preferences is an extensive one. “The algorithms are quite complicated but prove that statistics has a lot of practical uses,” he says.

Ved also works closely with the Seagate application that gives the recommendations to the engineers. These recommendations are collections of data-based reports and, with the abundance of reports, this becomes a very cumbersome process that the interns help alleviate. “We recommend new reports in the hopes it will give them the strategic intelligence and unique insight they didn’t have earlier,” Ved says. This project makes their day-to-day work easier and improves the company’s functionality as a whole.     

Campus and Community

Along with interning with Seagate, Ved is also an officer of STATCOM, a School of Statistics student group that does consultation work for nonprofits in the community. In previous consultations, STATCOM has worked with nonprofits like Nice Ride MN and HomeLine MN.  

Nice Ride MN places bicycle stations throughout the Twin Cities area for tourists, citizens, and whoever else wishes to rent them. STATCOM helped them learn more about their less frequent or walk-up customers. “The point was to help them achieve their goal of providing the best biking experience here in the Twin Cities,” Ved says. 

In April, STATCOM held a data competition for graduate and undergraduate University of Minnesota students. They used a public dataset, and after a few days working on it, they presented their findings to professors within the department and received insightful feedback. “It was a very meaningful contribution to our student group because it allowed us to learn practical skills,” Ved notes.

Through the School of Statistics, Ved also had the opportunity to participate in the Summer Statistical Collaboration Program. Students are paired with researchers on campus looking for statistical help on their projects. Ved was paired with Digital Arts, Sciences, and Humanities (DASH) program director Benjamin Wiggins and political science professor Timothy Johnson, both in the process of digitizing Supreme Court justices’ handwritten notes from the weekly conference meetings in which they would cast their votes. “These [we]re some of the most secretive meetings in the US with no record except for these handwritten notes,” Ved says. 

You can help transcribe the handwritten conference notes left by Supreme Court justices. Visit the SCOTUS Notes project website.  

Because most of the writing is illegible, researchers took pictures of the notes and are crowdsourcing them. Citizen scientists can look at the pictures from anywhere across the globe and help construct a translation of the notes. “We used natural language processing techniques, statistical algorithms, many machine learning techniques, and collaboration with language departments too,” Ved adds, acknowledging the importance of interdisciplinary collaboration. 

In his final year of graduate school, Ved feels ready for whatever the future may bring. He has applied for PhD programs to begin in fall 2018, including the University of Minnesota’s.

“I hope to do quality academic studies to further explore my research interests in a statistics PhD program,” Ved says. “I’ve experienced that the more you learn, the more seamlessly you can carry out your job.” He plans to continue working with Seagate until late July of this summer.

This story was written by an undergraduate student in CLA.

Share on: