My current research mainly focuses on data-intensive continuous optimization problems and in particular their applications in machine learning. The primary challenge here is to develop techniques that can practically deal with problems with up to millions of variables while guaranteeing fast local convergence such that high training accuracy can be achieved within reasonable amount of time. In one of my recent projects where for a broad class of l1-regularization problems such as sparse logistic regression, sparse inverse covariance selection, etc., we have successfully developed an efficient solver LHAC that is able to, for example, solve SICS problems with hundreds of thousand variables to an accuracy of 1e-6 in less than 10 seconds on my laptop. I gave a talk about the algorithm in both New York Machine Learning Symposium and MOPTA conference. You can read more about it here and, if you would like, play with our C/C++ implementation on GitHub.
Prior to study at Lehigh, I earned my undergraduate degree in Control Science and Engineering from Zhejiang University in China. I was honored to be selected into the Chu KoChen Honors College in my first year there and had the chance to study and collaborate with elite students in Zhejiang University from different departments. For two consecutive summers after coming to Lehigh, I interned at Siemens Corporate Research, where I worked on various machine learning related projects. Visit my profile on Linkedin for more of my education background and industrial experiences.