Signature Program Demonstrates How UC San Diego Undergraduates Learn to Transform Data into Action

Article Content

Since 2020 when the Halicioğlu Data Science Institute (HDSI) at UC San Diego established its Senior Capstone Program, it has become the premier series for all fourth-year undergraduate data science majors at HDSI – a pioneering interdisciplinary institute advancing data science and AI education within the new School of Computing, Information and Data Sciences (SCIDS) at the university. The program requires all students to undergo a two-quarter research experience with a faculty member or industry engineer before graduating.

“Students wrangle with open-ended problems and produce concrete deliverables while learning how to present their work to a variety of audiences, all while being mentored in a small group setting,” said Suraj Rampure, former UC San Diego lecturer, who served as the most recent program coordinator before joining the University of Michigan as a member of the teaching faculty in July 2024. The program coordinator provides students with the tools they need to perform the responsibilities of the program, which include submission of a final paper, a project report, a public-facing poster/showcase about their research and a code write-up. During the past year, 240 seniors participated in the program.

“Seeing the students’ final output, and seeing how they grow from sophomore year to the end of their senior year is rewarding,” said Rampure. “We have direct influence on everyone who graduates from the program.”

Recent UC San Diego and Senior Capstone Program Alumnus Diego Zavalza, who currently works as a data engineer for Amazon, developed a project aimed at finding a way to accurately and more effectively detect and treat sepsis.

“Sepsis typically arises from having two or more diseases present within patients and causes organ failure. Aside from having one of the highest mortality rates, there is no one way to detect sepsis besides running many labs and tests. This makes it incredibly difficult to effectively detect and treat patients in a timely manner,” explained Zavalza. “Our goal was to create a new detection model that accurately detected and predicted sepsis within five to 25 hours so that doctors could treat and prevent sepsis from developing in patients.”

Under the mentorship of HDSI Lecturer Kyle Shannon, Zavalza said that he and his project group members – Colin Tran, Asif Mahdin and Zhuji Zhang – spoke to doctors and nurses who treated sepsis patients. The group also used medical datasets to build a Long Short-Term Memory (LSTM)  model, which is a type of recurrent neural network effective for sequential data, such as processing a patient’s medical history.

Assistant Professor at HDSI Ben Smarr described another student group project involving the design of an integrative dashboard for patients to be able to see all of their medical results, tests and notes in one place, in order to work with their doctors to identify health patterns over time.

“The students had to develop different methods for helping patients pull data from different sources such as wearables, hospitals, journals, etc. They then had to work to design both a database backend and a visually intuitive interface for the front end. They also designed a security layer to ensure only the patient or people they allow can see their dashboard,” Smarr said.

Dr. Michael Kurisu, a family medicine physician specializing in long-term care for rare and/or chronic conditions and a pioneer in both integrative medicine and the use of new data sources to support patient histories and patient communication and advocacy, uses the dashboard at the Measured Wellness Clinic in California. In addition to the physician, the clinic consists of lifestyle educators, health coaches and students who are well-versed with data.

“The clinic has hired several of the (Capstone program) students and is building on their design to improve clinical dashboards. In fact, Measured Wellness has partnered with Ziva Health, a health technology company, which has taken pieces of the design from the students and is creating a software product that physicians and patients are currently using,” said Dr. Kurisu, adding that about a dozen students worked on the project each year over two years. “The students came from HDSI and bioengineering at UC San Diego. Many of the students have chosen to go into careers in the health tech industry.”

UC San Diego Department of Philosophy and HDSI Professor David Danks, who conducts research at the intersection of machine learning, philosophy and cognitive science, examines the ethical, psychological and policy issues around AI and robotics across a range of sectors. He and Capstone program mentors asked students last year to reflect on the ethical implications of their projects. The reflections led program leaders to realize that they needed to enhance the program’s discussion around the ethics of practicing data science.

“I met with each team to talk this through and realized that some students didn’t understand the relevance of ethical considerations in their work,” said Danks. “Since we cannot abdicate our ethical and societal responsibilities as we practice data science, we need to think about the ethical and societal implications of our work. The Senior Capstone Program now provides stronger insights for students to see how their work can matter beyond the technical.”

The Senior Capstone Program was established by HDSI’s former Assistant Teaching Professor Aaron Fraenkel, who operationalized the original vision of HDSI Founding Director Rajesh Gupta. “The Capstone program is vital as it accelerates student growth through real-world problem-solving and close faculty interaction, while also building strong industry connections that enhance both recruitment and collaboration. This initiative is key to preparing our graduates with the skills and industry ties needed for success,” said Gupta, who serves as the interim dean of SCIDS.

SCIDS includes the San Diego Supercomputer Center in addition to HDSI. The school meets a critical need in modern society to transform data into actionable knowledge across disciplines to advance science and technology. Established July 18, 2024, it is UC San Diego’s 12th school – just the fourth to be added in the 21st century. SCIDS aims to improve the human condition by better understanding how data shapes society, and to prepare the next generation of highly skilled workers driving artificial intelligence advancements.

For more information about the Senior Capstone Program, please contact ddanks@ucsd.edu.

Learn more about research and education at UC San Diego in: Artificial Intelligence