Mengchu Li

I'm a Postdoctoral Research Fellow in the Department of Statistics, University of Warwick. I will join the School of Mathematics, University of Birmingham, as an Assistant Professor of Statistics in October.

Previously, I completed my PhD in Statistics at the University of Warwick in 2023, supervised by Professor Yi Yu. I obtained my Master's degree in Mathematical Statistics (Part III) from the University of Cambridge in 2019, and Bachelor's degree in Mathematics and Economics from Durham University in 2018.

Research Interests: change point analysis, high-dimensional statistics, robust statistics and differential privacy.

Email  /  Google Scholar

profile photo

News

  • (March 2024) I created this personal website; Our paper on Federated DP is on arXiv; I will start working with Professor Gareth Roberts as a postdoc from April 2024.

Publication and preprints

Federated Transfer Learning with Differential Privacy
M. Li, Y. Tian, Y. Feng, Y. Yu.
Preprint, 2024.

Robust mean change point testing in high-dimensional data with heavy tails
M. Li*, Y. Chen*, T. Wang, Y. Yu.
Preprint, 2023. (*equal contribution)

On robustness and local differential privacy
M. Li, T.B. Berrett, Y. Yu.
Annals of Statistics 51(2), 717-737, 2023.

Contributions to robustness, local differential privacy and change point analysis
M. Li
PhD Thesis, 2023. (Harrison Award for highly commended thesis)

Network change point localisation under local differential privacy
M. Li, T.B. Berrett, Y. Yu.
Advances in Neural Information Processing Systems 35 (NeurIPS), 2022.

Adversarially robust change point detection
M. Li, Y. Yu.
Advances in Neural Information Processing Systems 34 (NeurIPS), 2021.

Multisite educational trials: estimating the effect size and its confidence intervals
A. Singh, G. Uwimpuhwe, M. Li, J. Einbeck, S. Higgins, A. Kasim
International Journal of Research & Method in Education, 45(1), 18-38, 2021.

Teaching

I have led tutorials/labs for the following modules at Warwick:

Introduction to Mathematical Statistics, Linear Statistical Modelling with R, Statistical Laboratory.

Software

I contributed to the R package changepoints: A Collection of Change-Point Detection Methods.

Events

(Seminar talks) U of Sydney Business School, U of York, U of Leeds, U of Southampton

(Conferences and workshops) ICSDS 2024, Workshop on heterogeneous and distributed data 2024, CMStatistics 2023, Workshop on Statistical Analysis of Networks 2023, RSS Conference 2023, Workshop on Change Point Analysis 2023, NeurIPS 2022 (with Scholar award), ICORS 2022, IMS Annual Meeting 2022, ISNPS 2022, Structural Breaks and Shape Constraints workshop 2022, NeurIPS 2021, StatScale workshop 2021

Services

I have reviewed for Bernoulli, Biometrika, Journal of the American Statistical Association, Journal of Machine Learning Research, Journal of the Royal Statistical Society: Series B, NeurIPS and ICLR.

The source code for this website is adapted from here.