Tuan Ngo Nguyen

Tuan Ngo Nguyen

PhD Student
Office: GS 725
Interests: Deep Unsupervised, Semi-supervised, Domain Adaptation, Interpretability
Advisor: Dr. Kwang-Sung Jun


I grew up in Saigon, Vietnam, a lovely city with unique culture,  moderate weather with two seasons (sunny and rainy), and many good foods. 

I had not been exposed much to computer science till my last undergraduate semester, and I even had a slightly negative biased view of programming (it is boring) from high school. My interest and focus have always been on math since I was young. I like the logic and beauty in math, but I don't feel I can work on rigid math problems for a long time without knowing whether they will be applicable. I went to study quantitative finance for my undergraduate degree; I learned many exciting things, but I was not too fond of a few aspects of it. Fortuitously, I met my previous graduate advisor during my last semester, and he introduced me to NLP and ML. I was also inspired by him to do research. Then, I self-studied more about NLP, ML, and computer science for some time. Then I did my Master's in CS at UofO before coming to UofA. I like the logic, openness, and accessibility in CS; I like to work on the math of some CS problems; I also can see the potential short- and long-term outcomes of my works.

​​​​​Research Interests

I am interested in learning and building theoretical-based deep learning algorithms such as self-supervised, semi-supervised, and domain adaptation. 

I am also interested in building interpretable, robust, and generalizable deep learning algorithms.  

What long-term project do you want to work on?

I am interested in building an effective method to transform or map the underlying distribution or statistics of the primary data's outputs learned by a DL algorithm into a target distribution. This transformation or mapping method would significantly benefit many deep learning algorithms; for example, mapping distributions of parts and the whole images or text in self-supervised, mapping distributions of unlabeled and labeled data in semi-supervised, and mapping distributions of source and target data in domain adaptation. 

What do you enjoy most about your work?

I enjoy learning and creating new, potentially valuable theories and applications. 

What are your career goals?

My career goal is to be a good researcher in an innovative industrial lab. My passion is in learning and creating new, potentially valuable theories and applications.  

​​​​Tell us something interesting about yourself!

"I like sports, and I am decent in many sports. I enjoy running (regularly), playing soccer (once or twice a week), and playing tennis (occasionally). I like reading good papers (lol) and good books."