I am a fourth year OxWaSp student supervised by Adam Johansen (Warwick) and Arnaud Doucet (Oxford) working on non-standard interacting particle systems for Fredholm integral equations of the first kind.
Before joining OxWaSP, I studied Mathematical Engineering at Politecnico di Torino.
I am interested in computational statistics, particularly sequential Monte Carlo methods and interacting particle systems. My work focusses on developping sequential Monte Carlo algorithms and particle methods for the solution of a particular type of integral equations which find applications in image reconstruction and density deconvolution.
In the past, I worked on algebraic based sampling techniques for discrete distributions and their link with permutations.
Publications & Pre-prints
Crucinio F.R., Doucet A., Johansen, A.M. A Particle Method for Solving Fredholm Equations of the First Kind. arXiv pre-print: 2009.09974
Fontana R., and Crucinio F. R. Orbit-based conditional tests. A link between permutations and Markov bases. Journal of Statistical Planning and Inference 205:(23-33), 2020. [.pdf]
My poster Sequential Monte Carlo for Fredholm Equations of the First Kind won the Bronze Award at STEM for Britain 2019 Competition in March 2019 (picture below).
Photos copyright John Deehan Photography Ltd
- Term 1: ST116 Mathematical Techniques
- Sequential Monte Carlo for Fredholm Integral Equations of The First Kind, Algorithms & Computationally Intensive Inference seminars, 6 March 2020, University of Warwick, Coventry, UK.
- Sequential Monte Carlo for Fredholm Integral Equations of The First Kind, 10th International Workshop on Simulation and Statistics, September 2019, Salzburg, Austria.
- Speeding up algebraic-based sampling via permutations, 4th Conference of the International Society for Nonparametric Statistics, June 2018, Salerno, Italy.
- SMC for Fredholm Equations of the First Kind, BayesComp 2020, January 2020, Gainesville, Florida, US.
Email: F.Crucinio at warwick.ac.uk
Office: MSB 4.14