Hey there!
I’m Vincent Maladière, an enthusiast of computer science, applied mathematics, and data science.
I have developed open-source software and deployed machine learning models in healthcare and fintech companies, and co-authored a paper on survival analysis. I have also taught data science at École Polytechnique and Dauphine.
I work at Probabl as a co-founding ML engineer, where I develop the skrub and hazardous libraries. You can reach me at:
:probabl. team
Inria's scikit-learn teamI contributed to
I took part in this paper
Teaching Models To Survive: Proper Scoring Rule and Stochastic Optimization with Competing Risks
published in arxiv (2024)
We propose a new SOTA model for survival analysis and competing risk analysis, trained on a proper scoring rule and plugged into a stochastic boosting tree. It is also much faster to train than current alternatives.
Guides
These resources helped me grasp the theory of ML, the main use-cases in MLOps and how to contribute to open source projects like scikit-learn. I use these extensively for teaching.
Courses
I teach data science at Polytechnique and Dauphine.
You can find below courses content and exercices.
Conferences
I talked at the following events and conferences
Notebooks
Tips and tricks on computer vision, audio and geo-clustering.
