Yuqi Gu

Email: yuqi.gu@columbia.edu.
Address: Room 928 SSW, 1255 Amsterdam Avenue, New York, NY 10027

I am an Assistant Professor in the Department of Statistics at Columbia University. I am also a member of the Data Science Institute. Prior to joining Columbia in 2021, I spent a year as a postdoc at Duke University, mentored by David B. Dunson. In 2020 I received a Ph.D. in Statistics from the University of Michigan, advised by Gongjun Xu. In 2015 I received a B.S. in Mathematics from Tsinghua University. My first name can be pronounced as /ju:-tʃiː/. My name in Chinese is 顾雨琦.

My research revolves around unobserved latent structures in statistics, machine learning, psychometrics and other social and biomedical applications:

  • Deep generative models and representation learning with latent variables: I study identifiability and other essential properties inherent in deep nonlinear models and probabilistic graphical models. One goal is to propose more interpretable models and discover potential causal explanations.
  • High-dimensional statistics in the presence of latent structures: The high dimensionality and the latent structures pose double challenges to statistical analyses. I aim to develop rigorous estimation and statistical inference methods with theoretical guarantees. I am actively exploring spectral methods as one promising approach.
  • Latent variable modeling in psychometrics and beyond: I develop principled statistical methods and theory to model educational and psychological data with substantively meaningful latent traits such as skills, attitudes, etc. I am also interested in other applications in biomedical sciences.

My research is partially supported by NSF Grant DMS-2210796 (sole PI).

I am looking for motivated students with a solid statistics and/or programming background to work with me.