Yi Yu  于怡




If you are also a statistician, you may have not cited my papers yet, but you must have cited my name tons of times, when you need an observation Yi, i = 1, ..., n. :)

I am a Lecturer in Statistics (on leave) in the University of Bristol and also a Research Associate in the Statistical Laboratory, University of Cambridge, under the supervision of Professor Richard Samworth.  Previously, I was a graduate student of Professor Zhiliang Ying at Fudan University. 


Lent 2014 -- Time Series (Part III) -- Lecturing

Michaelmas 2013, 2014 -- Principles of Statistics (Part II) -- Supervising


My research interests are high-dimensional statistics, network studies, survival analysis and applications in brain imaging data.

Working Papers

  1. Estimating time-varying connectivity for resting state fMRI data

  Ivor Cribben and Y.

• Estimating whole brain dynamics using spectral clustering

  Ivor Cribben and Y.

  1. Recurrent event survival analysis in dynamic network data

  Tony Sit, Zhiliang Ying and Y.

  1. Fused community detection.

  Yang Feng, Richard J. Samworth and Y.

Publications and preprints

  1. How many communities are there? (2014) arXiv preprint. [pdf]

  Diego Franco Saldana, Y. and Yang Feng

  1. Consistent cross-validation for tuning parameter selection in high-dimensional variable

  selection. (2013) arXiv preprint. [pdf, R codes]

  Yang Feng and Y.

  1. A useful variant of the Davis--Kahan theorem for statisticians. (2015) Biometrika, 102, 315-323. [pdf]

  Y., Tengyao Wang and Richard J. Samworth

  1. Modified cross-validation for penalized high-dimensional linear regression models. (2013) Journal of Computational and Graphical Statistics, to appear. [pdf, R codes]

  Y. and Yang Feng

  1. Invited discussion of Large covariance estimation by thresholding principal orthogonal complements by Fan et al (2013). Journal of Royal Statistical Society, Series B, Vol. 75, Part 4, 656-658. [pdf]

  Y. and Richard J. Samworth

  1. Apple: Approximate Path for Penalized Likelihood Estimators (2013). Statistics and Computing, Vol. 24, Issue 5, 803-819 [pdf]

  Y. and Yang Feng

  1. Oracle inequalities for the Lasso in the Cox model (2013). The Annals of Statistics, Vol. 41, No. 3, 1142-1165. [pdf]

  Jian Huang, Tingni Sun, Zhiliang Ying, Y. and Cun-Hui Zhang


  1. High-dimensional variable selection in the Cox model. (2010) [pdf]



  1. NCP, an R package for network change point detection using spectral clustering (2015). [source file, manual]

  2. fcd, an R package for fused community detection (2013). Available from cran.

  3. FusedPCA, an R package for community detection via fused PCA (2013). Available from cran.

  4. APPLE, an R package for Approximate Path for Penalized Likelihood Estimators (2012). Available from cran.

Ph.D. Thesis

  1. Contributions to high-dimensional variable selection. Fudan University, June 2013.

Invited talks

  1. Estimating whole brain dynamics using spectral clustering. (19th European Young Statisticians Meeting, Prague, Sep. 2015)

  2. Confidence intervals for high-dimensional Cox model. (University of Oxford, May. 2015)

  3. A useful variant of the Daivs--Kahan theorem for statisticians. (Shanghai Center for Mathematical Sciences, Jun. 2014)

  4. Fused community detection. (University College London, Feb. 2014; Columbia University, New York University, Stony Brook University, Mar. 2014, ERCIM, Pisa, Italy, Dec. 2014)

  5. Community detection via fused PCA. (Oxford-Man Institute of Quantitative Finance, Sep. 2013)

  6. Consistent cross-validation in penalized high-dimensional variable selection. (University of Cambridge, Apr. 2013)

• An introduction to community detection. (University of Warwick, Feb. 2013)

  1. Apple: Approximate Path for Penalized Likelihood Estimators. (Columbia University, Apr. 2012)


Room D2.04

Statistical Laboratory, Centre for Mathematical Sciences

Wilberforce Road, Cambridge, CB30WB.

Email: y.yu@statslab.cam.ac.uk

Phone: +44 (0)1223 337949