I am currently a machine learning researcher at the University of Bergen (Norway) in Michoel's group. Before that, I was a PhD candidate in the machine learning group of the University of Bergen. I studied mathematics and computer science in Toulouse (France), with 2 complementary master's programs in High-Performance Computing, Big Data and Machine Learning. My preferred programming language is python, but I also did some development for example in javascript for a web-based application.
On my GitHub profile, you can see most of my research and teaching projects.
As a researcher, I develop machine learning methods for biogenetics and my current project's objective is to identify new treatment targets (proteins) in neuropsychiatric disorders using causal machine learning.
My main PhD project aims at clustering time-series data in the context of ensemble weather prediction. From there, I became more and more interested in research questions related to clustering in general, notably taking into account the uncertainty on the number of clusters (PersiGraph and GraphApp), cluster validity indices (PyCVI and ClusterExp) as well as the impact of different distance metrics on the training and evaluation of machine learning models.
While doing an internship at the Nansen Center in Bergen (Norway) in 2019, I also carried out research on machine learning applied to ocean inverse problems, using Self-Organising Maps and Hidden Markov Models to infer subsurface data from surface data (SubMAPP).
During my PhD, I have been lucky enough to be a teaching assistant in 2 machine learning courses, INF264: Introduction to Machine Learning (INF264 and Python Crash Course) and INF265: Deep Learning (INF265 and PyTorch Tutorials). There, I was responsible for the practical part of the courses and could then design many exercises (with solutions) and tutorials.
In addition, I have been a lecturer / course coordinator in introductory courses in programming and python, DIGI611: Algoritmer og programmering (DIGI611, note: resources in Norwegian).