You Lu knew that the Discovery Analytics Center would be a great fit for his research when he applied to the Virginia Tech Ph.D. program in computer science. “I had heard about Professor Bert Huang and knew that he and I were working in a similar research area — graphical models,” said Lu. “I felt I could learn a lot from him and I am so glad he agreed to be my advisor.”
In June, You attended the International Conference on Machine Learning (ICML) in Long Beach, California, and presented his collaborative research with Huang, “Structured Output Learning with Conditional Generative Flows,” at the Workshop on Invertible Neural Nets and Normalizing Flows.
Lu and Huang also collaborated on “Block Belief Propagation for Parameter Learning in Markov Random Fields,” a paper in proceedings at the AAAI Conference on Artificial Intelligence earlier this year.
Lu’s research at DAC combine his interests in both theoretical analysis and in applying models to practical problems. “To solve structured prediction problems, you need to not only design models that work well in practice but also theoretically analyze their performance,” Lu said.
“At DAC I have the opportunity to work with many talented and great researchers and I have access to very good computing resources that are valuable to my work,” Lu said.
He focuses on generalized supervised learning that involves predicting structured objects rather than a scalar. He also resorts to other techniques like variational inference, graphical models, and deep generative models to solve structured prediction problems.
Lu said his research is relevant to many real world applications in natural language processing, computer vision, network science, etc. “One common application is image completion, filling out the missing parts of corrupt images,” Lu said.
Lu received a bachelor’s degree in computer science from Jilin University in China and a master’s degree in computer science from the University of Colorado Boulder.
Projected to graduate in 2021, Lu plans to pursue a career in research as a university professor or a research scientist in a laboratory.