CS&E Ph.D. Graduate Wins 2022 Graduate School's Best Dissertation Award
“Vipin is the best mentor I have ever had,” Jia said. “His passion for research and focus on making a real impact in the world is inspiring to me. He continues to be an invaluable resource to discuss new ideas and I am continuing to learn from him to this day.”
Jia first joined the Ph.D. program in 2015 where he started working with Kumar mapping Southwest Asian countries to help combat deforestation. That initial project launched him on the path of leveraging machine learning in real-world applications.
Most of us benefit from machine learning everyday, from using facial recognition to open a phone, to asking Amazon’s Alexa and Apple’s Siri for help. These systems have complicated, large-scale models that need large data samples to inform the algorithm. That isn’t a problem for tech giants, but is a big hurdle for other industries and researchers that want to leverage machine learning in the physical world.
Machine learning has the potential to speed up scientific discovery in freshwater science, agriculture and other natural sciences. However, gathering data samples in these fields is more difficult, time consuming and expensive, leading to smaller, more micro-level data samples that are harder to generalize. Jia’s work focuses on bringing machine learning into real-life applications by using a hybrid framework known as knowledge-guided machine learning (also referred to as physics-guided machine learning). This interdisciplinary approach blends principles of machine learning, which require larger data sets to make generalized predictions, with physics-based models that use smaller data sets to make predictions in a specific context.
This unique work requires collaboration with a number of different fields of study and gets the team out from behind a computer into some of the real-world labs on land and water. Jia has collaborated with aquatic scientists from the U.S. Geological Survey’s branch in Madison, Wisconsin, to predict water temperatures in different lake systems in the Midwest. He also worked with agriculture experts in the University of Minnesota’s College of Food, Agriculture, and Natural Resource Sciences to map primary crops in the United States, and predict crop yields and greenhouse gas emissions. Their work has made huge strides in finding commonalities in the physical, mathematical models in aquatic science, climate science and agriculture, which will continue to improve their machine learning models for other real-world problems.
Jia is now an assistant professor and researcher at the University of Pittsburgh where he is continuing to develop new machine learning algorithms that help solve a broader range of scientific applications, including mechanical engineering and fluid dynamics. With more time and practice in diverse fields, their models will continue to improve predictions and hopefully contribute to future scientific discoveries.
“My advice to other Ph.D. students in computer science is to push yourself to go beyond theories and make sure your work has real-world impact,” Jia said. “That is something Vipin stressed and it helped me so much with this research when I was literally getting my hands dirty working in these natural science fields. You need to take the initiative to implement, experiment and test out your programs in the real world.”
The son of two faculty members in China, Jia says that his parents' hard work and passion for research is what first inspired him to pursue this line of work. Coupled with the community of fellow researchers and faculty in the Computer Science and Engineering Department, Jia had everything he needed to push computer science into practical problem solving.
I am happy to see this work being recognized and I hope it will inspire others to apply computer science to solve bigger problems in the world,” he said. “I am especially proud that I won this award while at the University of Minnesota, because it was the most important time of my academic career. I loved my experience and it has shaped where I am heading in the future.”