Professor of Computer Science
I became a computer scientist by accident! While as a math student in China, I was introduced to Matrix Computation by an elegant and enlightening book by G. W. Stewart, which led me and prepared me to read a great monograph on Algebraic Eigenvalue Problems by J. H. Wilkinson, who passed away not long before I got a copy of his book. I sent my application to only one graduate school in America, the one at University of Maryland where Professor Stewart worked, and he still works there. I was accepted. I arrived at the College Park campus a week after the semester started. Following the address on the admission letter, I found myself in the Office of Graduate Study at the Department of Computer Science, not at the Math Department as I had intended. I was at a total loss in how to enroll myself into any class. For example, I knew the English words 'Operating' and 'Systems' separately, but I had not the slightest clue to what they meant when put together. That was the very first day on my journey to become a computer scientist.
I have since been designing and developing better and faster algorithms for matrix computations on evolving computers for solving many interesting problems, exploiting both mathematical properties and architectural characteristics. Many of the application problems I am interested in, especially in recent years, are related to signal and image processing, or data analysis and processing, or the new area of graph information processing. With my colleagues and students, I also worked on modeling, analyzing or estimating the performance and behaviors of large computer systems, using the computer systems themselves and, of course, my matrix tricks as well.
Return to CS@DUKE WOMEN page