Gianluca
Finocchio

Research

​

I am a Postdoc in statistics at the University of Vienna. 

My main research areas are: dimensionality reduction for ill-posed generalized linear models; Bayesian nonparametric inference; robust statistics.

My interests lie in the field of high-dimensional statistics, with a focus on the mathematical theory of machine learning methods.

 

Preprints

 ​

• An extended latent factor framework for ill-posed linear regression
  With Tatyana Krivobokova
  Preprint (2023) - arXiv:2307.08377
   

• Robust-to-outliers square-root LASSO, simultaneous inference with a MOM approach
  With Alexis Derumigny and Katharina Proksch
  Preprint (2021) - arXiv:2103.10420
   

 

Publications

 

• Posterior contraction for deep Gaussian process priors
  With Johannes Schmidt-Hieber
  Journal of Machine Learning Research 24.66 (2023): 1-49
   

• Two perspectives on high-dimensional estimation problems: posterior contraction and median-of-means
  PhD Thesis (2021)
   

• Bayesian variance estimation in the Gaussian sequence model with partial information on the means
  With Johannes Schmidt-Hieber
  Electronic Journal of Statistics 14(1): 239-271 (2020)
   

• Random matrix theory: Dyson Brownian motion
  Snapshots of modern mathematics from Oberwolfach (2020)
   

• SciSports: Learning football kinematics through two-dimensional tracking data
  With SciSport collaboration
  SWI 2018 Proceedings
  
   

​Invited & Contributed
  ​

• London, UK, Aug 22 – 25, 2023, 25th International Conference on CompStat
  Presented "Posterior contraction for deep Gaussian process priors"
   

• 34th European Meeting of Statisticians, Warsaw, Poland, Jul 3 – 7, 2023

  Presented "Iterative regularization methods for ill-posed generalized linear models"
   

• 2nd Workshop on High-Dimensional Data Analysis, Madrid, Spain, Mar 2 – 3, 2023
  Presented "Iterative regularization methods for ill-posed generalized linear models"
   

• 15th International Conference of the ERCIM WG, London, UK, Dec 2 – 3, 2022
  Presented "Iterative regularization methods for ill-posed generalized linear models"
   

• 16th World Meeting of the ISBA, Montreal, Canada, Jun 26 – Jul 1, 2022
  Presented "Posterior contraction for deep Gaussian process priors"
   

• 11th ICSA International Conference, Hangzhou, China, Dec 20 – 22, 2019
  Presented "Bayesian variance estimation in the Gaussian sequence model withpartial information on the means"
   

 

Workshops    

​

• Study group mathematics with industry SWI, Eindhoven, Jan 29–Feb 2, 2018
  Collaboration on the SciSport problem "Player evaluation in soccer using 2D tracking data"
   

​

Talks & Posters
​

• 27th Meeting of PhD students in Stochastics, Hilversum, May 27–29, 2019
  Presented PhD research
   

• The (International) Bayes Club, Amsterdam, Feb 2, 2019
  Presented talk "Bayesian variance estimation in Gaussian models with partial information"
   

• 26th Meeting of PhD students in Stochastics, Vught, May 28–30, 2018.
  Presented PhD research
   

• Oberwolfach Seminar: Statistical Inference for Complex Data, Oberwolfach, Germany, May 20–26, 2018
  Presented work related to bachelor thesis and Dyson Brownian motion
   

• Spring School "Structural Inference", Lubbenau, Germany, Mar 4–9, 2018
  Presented poster "Bayesian variance estimation in the Neyman-Scott problem"
   

 

Teaching

​

• Tutor, Statistics and Probability module for IBA at University of Twente, 2019–2020
  Frontal exercise sessions, exam supervision and grading
   

• Teaching assistant, Mathematical Statistics at MI Leiden University, 2017–2018
  Office hours, homework grading, exam supervision and grading
   

 

Academic experience & Education

​

• PhD candidate, Leiden University and University of Twente, 2017–2021
  Supervisors (Twente): Johannes Schmidt-Hieber, Katharina Proksch, Alexis Derumigny
  Supervisor (Leiden): Johannes Schmidt-Hieber
  Thesis: Two perspectives on high-dimensional models: posterior contraction and scale-invariant methods
   

• Master’s degree in Mathematics, University of Pisa, 2014–2017
  Grade:110/110 cum laude. Supervisors: Franco Flandoli, Dario Trevisan
  Topic: seminal paper by Karl Oeschlager on dynamics of systems of moderately interacting particles
   

• Bachelor’s degree in Mathematics, University of Pisa, 2010–2014 Grade:100/110. Supervisor: Marco Romito
  Topic: Dyson Brownian Motion
   

 

Scholarships

 

• Sergio Marchionne Student Achievement Awards, 2018
   

• Premi e Borse di Studio FIAT, 2011