Talks & Seminars

I’ll present the FairGGPR, a multi-criteria fair Gaussian regressor for insurance pricing, focusing on how different notions of fairness can be incorporated into actuarial predictive models.

A few days in Québec City 🍁 I had the chance to present my research on the FairGGPR and take part in the panel discussion “Methodological Challenges of Fairness in Insurance”.

My second seminar in Montréal! I was invited to present FairGGPR at the CRM Applied Mathematics Seminar at McGill University. I gave more or less the same talk I had presented at Concordia two weeks earlier, but this time I went a little deeper into the mathematical side of things.

First presentation of my postdoc! After defending my PhD on 20 August and arriving in Montréal two weeks later, I gave my first one-hour research seminar at Concordia University as part of the CANSSI Quebec Postdoc Day. I had the chance to meet other postdoctoral researchers along the way. It was also the first time I had the opportunity to properly present the Bayesian mathematics behind FairGGPR, rather than having to compress the methodology into a few slides and move on to the results.

On 20 August 2026, I publicly defended my PhD thesis in actuarial science at UCLouvain. My thesis explored statistical and machine learning methods for non-life insurance, with a particular focus on Bayesian modelling, predictive uncertainty, fairness, and interpretability.

The final conference of my PhD. A special shoutout to the KU Leuven team for letting me third-wheel the team for pretty much the entire PhD. I actually arrived in Korea a little early and spent some time on Jeju Island, where I hiked up Hallasan. A very fitting way to close this chapter.

Back to Stockholm, and back to my first research project on fairness — this time presenting it after its publication in the North American Actuarial Journal. This was probably my best presentation to date. It was also my first conference with my PhD supervisor, Prof. Donatien Hainaut!

A bit of a full-circle moment! This time, the ASTeRISK EOS Workshop was held at UCLouvain,co-organised with my PhD supervisor, Prof. Donatien Hainaut. We invited Mathias Lindholm (Stockholm University) and Yevhen Havrylenko (UNIL), and of course had the KU Leuven team back with us as well. It was really nice to see everyone together in Louvain-la-Neuve after all the workshops, and conferences over the previous years.

My first seminar! I was invited by Peter Hieber — after failing to keep my promise to stop by Lausanne during my summer trek through France, Italy, and Switzerland — to present my work on fairness in insurance pricing at the Actuarial Seminar at the University of Lausanne. The project was still in its early stages at the time, so it was also a fun opportunity to share some ideas that were very much still taking shape. It was a nice change from conferences and workshops, with a bit more time to dive into the research and discuss it properly.

Back to the Young Researcher Day at UCLouvain, almost exactly two years after presenting my second PhD project there for the first time. This time, I presented my final PhD project: FairGGPR, a multi-criteria fair Gaussian regressor for insurance pricing. It was also the first time I introduced its unofficial nickname, the Pesh Mel’ Bush model.

A very special one to close this conference season: the whole KU Leuven team was there! The conference itself was top tier — Tartu’s university campus is seriously impressive (let’s just not talk about the mosquitoes).

Back to IDS, this time in London! This was probably my favourite session I have ever been part of, presenting my work on fairness alongside Marie-Pier Côté and Mathias Lindholm. It was a great opportunity to exchange ideas with two researchers whose work I really enjoyed, and the discussion during the session was particularly memorable. Katrien Antonio captured the moment perfectly with a photo of the three of us afterwards!

Probably the most memorable conference of my PhD.

This was the first time I presented my work on fairness in insurance pricing — a project I had started towards the end of my second year of PhD and that turned out to be quite a challenging one to get into! I initially explored causal approaches before eventually focusing on demographic parity. I was not feeling particularly confident about this presentation, and my slides were… very much a work in progress.

One of my favourite conferences during my PhD! The Insurance Data Science Conference brings together two parallel perspectives, with an academic track and an industry track, which made it especially interesting to meet and exchange with people working in insurance outside academia. I really enjoyed the mix of research and industry discussions, and Stockholm itself was definitely a highlight — I may have slightly fallen for the Stockholm archipelago! I don’t have any photos from the conference itself, but I have plenty from the weekend kayaking trip around the archipelago afterwards.

This was another opportunity to present my work on synthetic insurance data to a broader data science audience. It was also nice to step a little outside the usual actuarial crowd and discuss how statistical and machine learning methods can be used to generate realistic synthetic insurance datasets.

About a year into my PhD, I presented my second PhD project for the first time at the Young Researcher Day organised by my department at UCLouvain. The event is a great opportunity to share our research with fellow colleagues in a more friendly atmosphere. I have particularly good memories of this presentation — it was probably the first time I really felt confident presenting my work. I introduced my new project on generating synthetic insurance data, which is also where my interest in Bayesian statistics really started to grow.

Ten months into my PhD, this was my first presentation in front of a truly international audience (a little more academic pressure). I travelled to Edinburgh with Rodrigue Kazzi, who actually submitted the final version of his PhD thesis at the airport when we landed!

My second conference presentation, this time back home in Belgium. I presented my work to colleagues and professors from my department, as well as researchers from UNamur. It was a nice opportunity to share my research with a slightly different audience and to present how clustering methods can be used to explore structure and heterogeneity in insurance portfolios.

This was my first conference presentation, just four months into my PhD. I attended PARTY alongside my former colleagues Vanessa Hannah and Keivan Diakite. It was also where I first met members of the KU Leuven team. I presented my first research talk on clustering methods for heterogeneous insurance datasets and received the Best Presentation IAALS Prize. PARTY is a workshop organised every two years by Séverine Arnold (UNIL) and Corina Constantinescu (University of Liverpool), bringing together young researchers working on actuarial risk.