HDSI Faculty Feature Article Series: I Am Data Science
with Assistant Professor Rose Yu, Ph.D.
By Trista Sobeck & Bobby Gordon
Rose Yu, Ph.D. is an assistant professor at UC San Diego in the Halıcıoğlu Data Science Institute and the Department of Computer Science and Engineering. Her work and interests lie in AI, machine learning (ML) for large-scale spatiotemporal data. She develops algorithms and models to extract insights from studies the behavior and patterns of these types of data.
Dr. Yu employs a largely accessible definition of data science by explaining that it is, “a systematic study of data to build and organize knowledge of our world.” Her definition is so simple, yet all-encompassing as it calls attention to what all of us try to do daily—build and organize the knowledge we have.
Since our knowledge is continually evolving and growing, the field of data science continues to grow with us. “Data science is a highly interdisciplinary field which covers the full spectrum from system design, model development, theoretical analysis to interacting with domain scientists,” she clarifies. Dr. Yu explains that when you’re a member of the data science field, you must constantly cross domain boundaries and learn knowledge in other areas.
“My long-term interest is to accelerate science and engineering applications,” she says. “With rapid advancements in sensing devices and measurement instruments, we now have access to a massive amount of spatiotemporal data,” she explains. She continues to explain that her role with HDSI is to develop machine learning methods to help enable scientific discovery.
Dr. Yu’s Lab
Currently, Dr. Yu is tackling data science challenges in a wide range of topics. From climate modeling to traffic prediction, to disease forecasting. She encourages future data scientists to look for complex information that is inhibiting progress to occur.
“Search for data science challenges in your own field where the volume and complexity of data are becoming a key bottleneck for the field to progress,” she says. If you can brainstorm how to formulate the domain problem as a data science problem, information can become a bit clearer and simpler.
Currently, Dr. Yu is examining physics-guided AI and designing hybrid approaches to integrate first principles with data-driven inference. “In this way, we can combine domain knowledge [with] science/engineering disciplines, guarantee the physical consistency of ML predictions, and contribute towards trustworthy AI in science,” she explains.
Dr. Yu’s prediction? “Data science will greatly accelerate the development of science and engineering. It will provide principled tools that have a great impact far beyond.” Here’s to a truly bright future.
Recently recognized with a 2021 Facebook Research Award, Dr. Yu’s work also encompasses the DeepGLEAM COVID-19 forecasting model which combines the signal of a discrete stochastic epidemic computational model GLEAM (Global Epidemic and Mobility Model) with a deep learning spatiotemporal forecasting framework to further improve predictions to study the spatiotemporal COVID-19 spread. The hybrid model leverages rich real-world data about COVID-19: when a person has been infected; where they have traveled; death records; travel and re-opening restrictions from all 50 states. DeepGLEAM is a part of the national COVID-19 ensemble forecast project with the CDC (Centers for Disease Control).
Connect with Rose Yu
Dr. Yu’s Homepage: https://roseyu.com
HDSI/UC San Diego Faculty Profile: https://datascience.ucsd.edu/about/contact/directory/name/rose-yu/