Qi Xu

I am a tenure-track assistant professor in the School of Statistics at the University of Minnesota. Prior to this, I was a postdoctoral researcher in the Department of Statistics & Data Science at Carnegie Mellon University, working with Kathryn Roeder and Jing Lei. Previously, I received my PhD from the University of California, Irvine, advised by Annie Qu. Before that, I earned an M.S. in Statistics from the University of Illinois Urbana-Champaign and a B.S. in Mathematics from Tongji University.

I am recently interested in:

  • How can we integrate heterogeneous datasets for prediction, estimation, and inference with statistical rigor? [method/theory] [ 1, 2, 3, 4, 5, 6, 7 ]
  • How can we integrate predictive and generative AI models into statistical problems? [method/theory] [1]
  • How will AI change the research pipeline and community? [application]

Email me at qixu[at]umn.edu. See my full publication list in Google Scholar .

Recent Updates

Aug 04, 2026 I will give a talk at the JSM Biometrics early career award session [slides].
Jun 14, 2026 I will give a talk on semiparametric inference under blockwise missingness at the ICSA Applied Statistics Symposium [slides].
Jun 09, 2026 Our paper representation retrieval learning for heterogeneous data integration [Arxiv] is accepted in Journal of the American Statistical Association. We propose a sparse dictionary learning type method for multi-task learning, incorporating blockwise missing covariates. Our main finding is that the generalization error is governed by the sum of square-root of integrativeness of each representer in the dictionary. A selective integration penalty is proposed to encourage integrativeness of representers to improve prediction performance.
Feb 05, 2026 New Preprint Available: sparse group principal component analysis [Arxiv]. This work proposes an efficient iterative algorithm with double thresholding steps to estimate principal components where data is collected from multi-view, especially multi-cell type in gene expression data.
Jan 11, 2026 Our paper Blockwise Missingness meets AI [ArXiv] just won the ASA Biometrics Early Career Award.