Raphaël Reynouard

Postdoc Researcher · Machine Learning · Generative AI · Model Checking
EAWAG · Dübendorf, Switzerland

I am a researcher specialising in machine learning, with a PhD focused on Markov models. I have extensive experience in Python, demonstrated through the development of the Jajapy library. My recent postdoc at Eawag involved applying deep learning methods to hydrology, where I acquired strong skills in PyTorch. I bring a robust background in statistical modelling, algorithm development, and practical data analysis.

Skills

Programming Languages & Tools
  • python
  • pytorch
  • numpy
  • scikit
  • r
  • java
  • latex
  • overleaf
  • arduino
  • linux
  • raspberry_pi
Technical Knowledge
  • Statistical Foundations of Machine Learning
  • Automata Theory & Markov Models
  • Deep Learning
  • Model Checking
  • Algorithm Design
  • Data Analysis and Visualization
  • Numerical Methods
  • Physics-Constrained Modeling
Languages
  • French, native
  • English, full professional profeciency
  • German, basics

Experience

Postdoc Researcher

EAWAG · Dübendorf, Switzerland
  • Deep Learning · Generative AI · Hydrology
  • Developed deep learning and generative AI methods for structure learning and parameter estimation of mass-conserving process-based models.
January 2024 - August 2024

IT Specialist

Expo Kilim · Brussels, Belgium
  • Database management
  • Website maintenance
  • CCTV installation and set up
June 2018 - September 2019

Education

Ph.D. in Computer Science

Reykjavík University · Reykjavík, Iceland
  • Machine Learning · Model Checking
  • Anna Ingólfsdóttir · Giovanni Bacci
  • On learning stochastic models: from theory to practice
  • Developed and implemented unsupervised machine learning methods for parameter estimation of Markov models.
August 2020 - November 2023

M.S. in Computer Science

Université Libre de Bruxelles · Brussels, Belgium
September 2018 - June 2020

B.S. in Computer Science

Université Libre de Bruxelles · Brussels, Belgium
September 2015 - June 2018

Publications

On learning stochastic models: from theory to practice

R. Reynouard
PhD Thesis

A MM Algorithm to Estimate Parameters in Continuous-time Markov Chains

G. Bacci, A. Ingólfsdóttir, K. G. Larsen, R. Reynouard
QEST'23

Jajapy: a learning library for stochastic models

R. Reynouard, G. Bacci, A. Ingólfsdóttir
QEST'23

Active Learning of Markov Decision Processes using Baum-Welch algorithm

G. Bacci, A. Ingólfsdóttir, K. G. Larsen, R. Reynouard
ICMLA'21

Online Learning of non-Markovian Reward Models

G. Rens, J.-F. Raskin, R. Reynouard, G. Marra
ICAART'21

Tool

Jajapy

A Machine Learning Python library for Markov models
A python library implementing the Baum-Welch algorithm on various kinds of Markov models.
  • Learning HMMs, MCs, MDPs and CTMCs from traces.
  • Parameter estimation for synchronous composition of CTMCs.
  • Parameter estimation for PCTMCs.
  • Compatibility with Prism and Storm.
Since 2022

Research activities

Artifact evaluation committee member

QEST'23

Artifact evaluation committee member

QEST+FORMATS'24

People have been honored to have work with

Side Projects

Playground

Visualization of several algorithms that facinate me

Which is AI

Guess which of the pictures is AI generated

files