Avatar

Nina Deliu

Continuous Stats Learner, Piecewise Teacher.

MEMOTEF, Sapienza University

MRC-BSU, University of Cambridge

Welcome!

I am an Assistant Professor (RTDA) in Statistics at the MEMOTEF Department of Sapienza University of Rome (IT), and a long-term visitor at the MRC - Biostatistics Unit, University of Cambridge (UK), where I was a Postdoc in 2020 — 2021.

I collaborate with different academic (and other research) institutions worldwide, some of which I visited during my PhD. Active international collaborations include University of Cambridge, National University of Singapore and University of Toronto, where I am also part of the IAI Lab. In Italy, I am leading some work in collaboration with ISTAT and NADO Italia. I am in the editorial board of YoungStatS, the blog of Young Statisticians Europe (YSE), and curate the Women in Statistics and Data Science Twitter account. I am also a member of ISCB, IMS, ACM and SIS.

I embrace diverse, and perhaps too many to mention, areas of statistics. My primary research interest focuses on sequential decision-making problems, intersecting areas of statistical inference, Bayesian statistics, reinforcement learning (RL) & multi-armed bandits (MABs), and adaptive designs / experiments. In the specific, the following two research questions guide my work:

  1. How can we use/develop MABs and RL algorithms for adaptive experiments that may solve practical problems in areas such as clinical trials, mobile-Health, education and other behavioral sciences? and;

  2. How and when can we perform valid inference in adaptively collected data?

More recently, I became interested in copula models for representing data dependencies, and how these can be used for constructing highest density regions.

My research is inspired by the numerous challenges arising in real-life applications and directed towards providing concrete benefits in these areas. I strongly believe that the theoretical and methodological progress should go along with the concrete real-world needs, and be not simply good, but also good for something.

Education

  • PhD in Methodological Statistics, 2021

    University of Rome La Sapienza

  • Double Master Degree in Bayesian Statistics and Decision Sciences, 2017

    Universitè Paris Dauphine & University of Rome La Sapienza

Interests

  • STAT Sequential Decision-Making, Inference, Bayesian Analysis, Copulae
  • ML Reinforcement Learning, Multi-armed Bandits
  • APP Biostat, Design of Experiments, Education

Latest Publications