$\mathbb{P}$robably Approximately Wrong

An infrequent blog, by Nicola Branchini

About me

Link to my resume
Slides about some of my recent work

Hi! I am Nicola Branchini, a PhD student in Statistics in the School of Mathematics at the University of Edinburgh, advised by Prof. Víctor Elvira. Sometimes I blog as well here. I am also an ELLIS PhD student, working with and co-supervised by Prof. Aki Vehtari in Aalto University.

My real interests are broad, spanning computational statistics and statistical/probabilistic machine learning, with a focus on methodology. For my PhD, I have been focussing on developing methodology in Monte Carlo (MC), with a particular focus on importance sampling (IS). The concept of IS is all about “sample efficient” - in ML terminology - MC integration, and of understanding MC integration as optimization over probability densities. This is relevant in a number of applications beyond “just” Bayesian computation.

I like collaborating with people. Feel free to drop me an email (and to ping me again if I do not reply).

News

- Contributed talk accepted in a minisymposium at the 14th international conference on Monte Carlo methods and applications (MCM) 2023, on “Generalized Self Normalized Importance Sampling”. ✰
- Causal Entropy Optimization to appear in AISTATS 2023. ✰
- Received ISBA Travel Award for BayesComp 2023.
- Accepted for attending the 2023 Probabilistic Numerics School in Tübingen
- Poster on “Generalized Self Normalized Importance Sampling” accepted at BayesComp 2023.
- I received the Turing Enrichment Scheme (Placement Award, for 6 months) offer.

Reviewing

Journals

Statistics and Computing, Statistics and Probability Letters

Conferences

AISTATS 2023, AABI (workshop) 2023, NeurIPS 2023, ICLR 2024, AISTATS 2024, NeurIPS workshop on Bayesian decision making and uncertainty 2024, AISTATS 2025

Talks & Posters

Awards

"Basically, I'm not interested in doing research and I never have been. I'm interested in understanding, which is quite a different thing. And often to understand something you have to work it out yourself because no one else has done it"

- David Blackwell

"Getting numbers is easy; getting numbers you can trust is hard."

- Ron Kohavi, Diane Tang, Ysa Xu (from the book "Trustworthy Online Controlled Experiments")

Some background

Previously, I was a Research Assistant at the Alan Turing Institute, working within the Warwick Machine Learning Group and supervised by Prof. Theo Damoulas. Previous to that, I was a Master’s student in the School of Informatics at the University of Edinburgh where I was supervised by Prof. Víctor Elvira working on auxiliary particle filters. As undergrad, I studied Computer Science at the University of Warwick, where I did my BSc dissertation on reproducing AlphaZero supervised by Dr. Paolo Turrini.

Random selection of nice reads

Worth having the physical version.

Blogroll