Khoa Doan, DAC Ph.D. student in computer science

After graduating with a bachelor’s degree in computer science from Webster University, Khoa Doan entered the workforce. For the next few years, he held positions as a software developer and data engineer in the advertising industry and at NASA and gained experience processing large datasets.

“I came to appreciate both the theoretical and practical contributions,” said Doan. “Working with large datasets is tricky because solutions become much more constrained. The challenge is what interests me and it makes me really happy if I am able to solve a problem.”

Doan decided to pursue a master’s degree in computer science at the University of Maryland. When looking for a Ph.D. program, he was attracted to Virginia Tech and the Discovery Analytics Center “because of a good mix of strong theoretical foundation and practical research objectives. There is a diversity and plethora of research opportunities, especially in things that matter.”

And since being at DAC, he said, “I have learned a lot from the DAC community. I have good friends with both similar and diverse research interests.”

Doan’s main research interest is in scalable machine learning and data mining. His current focus is on deep hashing for similarity search, using neural networks as a basis for efficiently “searching” for similar items in very large databases. For example, he searches for similar documents in news articles, books or papers, and images.

“This problem is very hard because we have to pay attention to both efficiency — how to retrieve the items fast, and sometimes in real-time — and effectiveness,” said Doan. “Items can be similar because of similar words, but also because of similar authors, or similar topics, thus it is very difficult to choose the right concept to describe similarity and convert these informal concepts into mathematical equations.” Doan is also working with his advisor, Chandan Reddy, on research with Criteo, a leading advertising company that has made a significant investment in machine learning.

“Having worked in the advertising industry, solving computational problems in this field is of interest to me, as well, and is a great opportunity,” Doan said.