Niccolo' Anceschi, Ph.D.

I am a Postdoctoral Associate in Statistics at Duke University in Durham, NC (USA).

My research focuses on the development of scalable Bayesian methods for interpretable latent structure discovery in high-dimensional, multimodal, and heterogeneous data, with emphasis on uncertainty quantification and efficient inference — including MCMC, variational methods, and expectation propagation — implemented in accessible open-source software.

Current and recent applied work includes collaborations with scientists at Merck Sharp & Dohme on multi-omics data integration for precision medicine, and with academic investigators in environmental health, toxicology, and ecology.
This includes application-specific methodology development and simulation-based power analysis for studies of toxicant-induced neurodegeneration, and statistical consulting for studies on air pollution-induced pulmonary inflammation and firefighters' exposure to harmful chemical mixtures.


Publications


Anceschi N., Rigon T., Zanella G. & Durante D. (2026) Optimal and computationally tractable lower bounds for logistic log-likelihoods, Accepted for Publication in Biometrika - preprint at arXiv:2410.10309

Anceschi N., Ferrari F., Mallick H. and Dunson D. (2026) Bayesian joint additive factor models for multiview learning, Accepted for Publication in Biometrics - preprint at arXiv:2406.00778

Poworoznek E., Anceschi N., Ferrari F. and Dunson D. (2025) Efficiently resolving rotational ambiguity in Bayesian matrix sampling with matching, Bayesian Analysis, Advance Publication, 1-22, (2025)

Anceschi N., Fasano A., Durante D. and Zanella G. (2023) Bayesian conjugacy in probit, tobit, multinomial probit and extensions: a review and new results, Journal of the American Statistical Association, 118 (542), 1451-1469

Anceschi N., Hidalgo J., Plata. C., Bellini T., Maritan A., Suweis S. (2019) Neutral and niche forces as drivers of species selection, Journal of Theoretical Biology 483, 109969


Preprints


Mauri L., Anceschi N., & Dunson D. (2026+) Spectral decomposition-assisted multi-study factor analysis, arXiv:2502.14600

Anceschi N., Fasano A., Franzolini A. and Rebaudo G. (2026+) Scalable expectation propagation for generalized linear models, arXiv:2407.02128


Full PDF versions of my CV and resume are available here.

Current Position

Postdoctoral Associate

Duke University, Department of Statistical Science, Durham (NC)
October 2022 - Present

Education

Ph.D. in Statistics

Bocconi University, Milan (IT)
September 2018 - January 2023

M.Sc. in Physics

University of Milan, Milan (IT)
October 2014 - April 2017

B.Sc. in Physics

University of Milan, Milan (IT)
October 2010 - April 2014

Work Experience

Pricing Analyst

June 2017 - June 2018

Teaching

A.Y. 2024-2025
DUKE UNIVERSITY, DURHAM, USA
The Mathematics of Regression - STA211 (B.Sc. Statistics) - Course Instructor

A.Y. 2021-2022
BOCCONI UNIVERSITY, MILAN, ITALY
Computer Science - Module 2 (Computing Theory and Algorithms) (B.Sc. in Mathematical and Computing Sciences for Artificial Intelligence) - Teaching Assistant

UNIVERSITA' DEGLI STUDI DI BERGAMO, BERGAMO, ITALY
Statistics (B.Sc. in Economics and Management) - Teaching Assistant

A.Y. 2020-2021
BOCCONI UNIVERSITY, MILAN, ITALY
Computer Science - Module 2 (Computing Theory and Algorithms) (B.Sc. in Mathematical and Computing Sciences for Artificial Intelligence ) - Teaching Assistant

A.Y. 2019-2020
BOCCONI UNIVERSITY, MILAN, ITALY
Theoretical Computer Science (B.Sc. in Economics, Management and Computer Science) - Teaching Assistant

Other Interests