“In biomedicine, the consequences of failed or biased AI decision-making can be fatal, and there are many privacy risks in existing Al algorithms… These challenges are what really keep me up at night.”
Dr. Xiaoxiao Li arrived at the UBC Electrical and Computer Engineering Department this August, from her previous position as a postdoc at Princeton University. As a new faculty member in ECE, this year she’ll be teaching ELEC 400M in Term 2. Her research focuses on developing advanced AI algorithms and trustworthy AI systems, especially for healthcare applications.
We spoke to her to learn about her career path, her outlook on the uses of AI and the future of this field, and her perspective on teaching.
What has your career path looked like so far?
Before joining the Electrical and Computer Engineering department at UBC, I spent a year at the Computer Science Department at Princeton University. I received my Ph.D. from Yale University in 2020. I received my Bachelor’s degree from Zhejiang University, during which I spent three years in the beautiful city of Hangzhou and one year at Harvard University.
Tell me about your field of study. What is your area of expertise?
I am an Artificial Intelligence (AI) researcher. My current research lies in developing advanced deep learning algorithms/theories and applying them to solve real problems in the field of healthcare. Our group’s research covers computer vision, natural language processing, trustworthy AI (privacy, explainability, robustness, and fairness), federated learning, medical image analysis, and health informatics.
Why did you choose this field? What excites you about the research and work you do?
I’m interested in designing next-generation AI algorithms that are more interpretable, efficient, accurate, and robust in healthcare applications. Unfortunately, AI has not been widely used in actual clinical practice. The primary reason is that current AI algorithms fail to explain how and why the AI makes decisions. In biomedicine, the consequences of failed or biased decision-making can be fatal. There are also many privacy risks with existing Al algorithms. So, when you think about it, solving these issues is critical for biomedical applications under ethics and regulation requirements. Those challenges are what really keep me up at night.
What do you think is in store in the future for your field of study?
I think we will begin to see a broader spectrum of algorithms that are specialized for different clinical applications, and which will focus on accuracy and solving issues with trustworthiness, such as explainability, fairness, and privacy. These next-generation AI algorithms will address issues such as learning from heterogeneous data, unlabeled data, insufficient data, and data distributed in noisy, dynamic, or adversarial environments. For applications in the realm of healthcare, these next-generation AIs will help us to better understand diseases- for example, what biomarkers and environmental factors are causing cancer and other complex diseases. So, the near future of medicine will see a significant impact as well.
What courses will you be teaching in the upcoming year? How would you describe your teaching style?
I will be teaching ‘ELEC 400M: Machine Learning Fundamentals for Engineers’ in term 2. In this course, I’ll emphasize both a theoretical and practical understanding of the materials. I hope that students can gain knowledge no matter whether they want to go into the industry or into academia. I really care about the needs of each student, and the most important ability that I want to teach is the skill of ‘learning to learn.’
You’ve moved to Vancouver recently. What do you think of UBC and Vancouver so far? Any places you are looking forward to visiting?
I really enjoy working at UBC and living in Vancouver so far. I’m waiting for wintertime so I can go skiing.
What are your interests outside of work?
During my spare time, I like playing sports (basketball, tennis, badminton, golf, yoga, hiking, skating- you name it!) I’m also into cooking, calligraphy, photography, and reading science fiction. If I have a vacation, I’ll take the opportunity to travel with my family.
