2016-lsa

Yuekai Sun

By | | No Comments

Yuekai Sun, PhD, is Assistant Professor in the department of Statistics at the University of Michigan, Ann Arbor.

Dr. Sun’s research is motivated by the challenges of analyzing massive data sets in data-driven science and engineering. I focus on statistical methodology for high-dimensional problems; i.e. problems where the number of unknown parameters is comparable to or exceeds the sample size. My recent work focuses on two problems that arise in learning from high-dimensional data (versus black-box approaches that do not yield insights into the underlying data-generation process). They are:
1. model selection and post-selection inference: discover the latent low-dimensional structure in high-dimensional data and perform inference on the learned structure;
2. distributed statistical computing: design scalable estimators and algorithms that avoid communication and minimize “passes” over the data.
A recurring theme in my work is exploiting the geometry of latent low-dimensional structure for statistical and computational gains. More broadly, I am interested in the geometric aspects of high-dimensional data analysis.

A visualization of an algorithm for making accurate recommendations from data that contain shared user accounts.

A visualization of an algorithm for making accurate recommendations from data that contain shared user accounts.

 

e0b4951f668bc44f55_l_0b66b

Vu-Minh Chieu

By | | No Comments

Vu-Minh Chieu is an assistant research scientist at the GRIP (Geometry, Reasoning, and Instructional Practices) laboratory, directed by Patricio Herbst at the School of Education. Chieu’s research focuses on personalized learning, intelligent tutoring systems, computer-based simulations, human–computer interaction, and technology-enhanced professional learning. Chieu’s research aims at describing and explaining how theories of learning and instruction can be applied to the design of learning technologies as well as how technologies can be designed to build, test, and refine theories of learning and instruction. Patricio Herbst and Chieu has designed and developed LessonSketch (www.lessonsketch.org), a virtual lab that supports practice-based learning for mathematics teacher development. Chieu and his colleagues has been developing computational models and tools to support the collection of learners’ interaction with LessonSketch and with each other (e.g., their use of resources and tools in LessonSketch, contents they create, and logs of their forum discussions), to analyze the collected data (e.g., analysis of correlations between their use of resources and the quality of their discussions), and to use the analyzed data to inform the design of learning technologies as well as to personalize learning experiences for learners (e.g., the design of software features to improve the quality of forum discussions).