Pierre Bellec

Pierre Bellec is an Associate Professor in the Department of Statistics at Rutgers University, where he has been a faculty member since 2016 and was promoted to Associate Professor with tenure in 2021. His office is located in Hill Center 406 at 110 Frelinghuysen Road, Piscataway, NJ 08854.

Dr. Bellec received his PhD from ENSAE ParisTech, France in 2016 under the supervision of Alexandre Tsybakov. Prior to that, he completed a Part III (MASt) at the University of Cambridge, UK in 2012 and earned his Diplôme d'Ingénieur from Ecole Polytechnique, France in 2011.

Dr. Bellec's research focuses on high-dimensional statistics, aggregation of estimators, shape constrained problems in statistics, and probability theory. His work has significant implications for machine learning and statistical inference in high-dimensional settings. He has made important contributions to regularization methods, M-estimators, and uncertainty quantification in complex statistical models. His recent publications show a strong focus on asymptotic theory, robust statistics, and high-dimensional inference, with novel methods for error estimation, adaptive tuning, and bias correction in regularized estimators.

Dr. Bellec has received several prestigious awards and honors, including:

• IMS Fellow (2023)
• NSF CAREER award DMS 1945428: "Post-Differentiation Inference" (2020-2024, $400,000)
• NSF award DMS 2413679: "Uncertainty quantification for iterative algorithms" (2024-2027, $225,000)
• NSF award DMS 1811976: "Uncertainty Quantification in High-Dimensional Structured Regression Problems" (2018-2022, $180,000)
• Blaise Pascal PhD Award (2017)

Dr. Bellec has advised several graduate students, including Takuya Koriyama (now at Chicago Booth), Yiwei Shen (now at Meta/Facebook), and Kai Tan (current PhD student). He has also mentored numerous undergraduate students through REU programs. He serves as an Associate Editor for the Annals of Statistics and has been involved in program committees for major conferences including the Conference on Learning Theory (COLT) and the Conference on Neural Information Processing Systems (NeurIPS).