Brief and baseline
Everything students received to get started in two days.
Lectures · Practical work · Material
Teaching
Slide decks, notebooks and datasets from the courses I teach, in one place. Free to reuse for teaching. A link back is appreciated.
Since 2020 · Lecturer
EPITA
Machine learning, numerical methods and programming for the MSc and the MSc in Artificial Intelligence Systems.
MSc Python & Algorithm Workshop
Seven workshops introducing programming through numerical problems: integer and floating-point arithmetic, matrices, linear and nonlinear systems, interpolation.
Workshop 7: Oral presentations
MSc Python
From the basics of the language to scientific computing, visualisation, dataframes and engineering tooling, with an insurance case study.
Why Python?
Read
Installing Python
Read
PW 2: Application
ExercisesNotebookUsual mistakes
PW 3: Scientific computing
LectureNotebookCorrection
PW 4: Data visualisation
LectureNotebookCorrection
PW 5: Data manipulation
LectureNotebookCorrection
PW 6: Engineering tools
Lecture
Additional exercises: basics
ExercisesNotebookCorrection
Additional exercises: exam-like
ExercisesNotebookCorrection
Case study: freMPL dataset
NotebookfreMPL.csvIris.csv
MSc AI Systems Numerical Algorithms
Optimisation for machine learning: calculus refresher, unconstrained and constrained optimisation, numerical linear algebra, ML applications.
Lecture 3: Constrained optimisation
SlidesApplication
Lecture 4: Numerical methods in linear algebra
SlidesPCA application
Lecture 5: Machine learning applications
Notebook
Lecture 6: Oral presentations (Titanic challenge)
MSc AI Systems Bayesian Machine Learning
Bayesian statistics, latent variable models and the EM algorithm, variational inference with NLP applications, and causal inference.
Lecture 2: Latent variable models and the EM algorithm
Lecture
Lecture 3: Variational inference and application to NLP
Lecture
PW 1: Naive Bayes classifier
Notebook
PW 2: GMM, probabilistic k-means and PCA
Notebook
PW 3: Topic modelling with LDA
Notebook
PW 4: Bayesian linear regression
NotebookConjugate priorsCorrection
Dataset: Defra consumption
CSV
Since 2019 · Lecturer, Data Science pour l'Actuaire (DSA)
Institut des Actuaires
Text mining, NLP and active learning for practising actuaries, plus machine-learning methods for individual mortality modelling.
2025 Machine learning for individual mortality modelling
Building mortality scores from individual data, and what algorithmic bias means in that setting.
Mortality scores
Notebook
Introduction to algorithmic bias
Lecture
Datasets
labeled.zipunlabeled.zip
2019 to 2021 Text mining and introduction to active learning
Preprocessing for NLP, vector representations, topic models, deep learning for text, and annotation strategies.
Introduction: Preprocessing in NLP
Lecture
Lecture 1: Vector representations and topic models
Lecture
Lecture 2: Deep learning for NLP
Lecture
Lecture 3: Introduction to active learning
LectureAnnotations
Animations: recurrent networks
RNN forwardRNN backpropBRNN forward
Practical work: Topic modelling
NotebookCorrection
Since 2024 · Lecturer, MSc Actuarial Science
Cnam
Six hours on fairness in insurance, integrated into the Contemporary Actuarial Science course.
MSc Actuarial Science Fairness in insurance
How discrimination arises in actuarial models, how to measure it, and what post-processing can and cannot fix.
Mortality score: hands-on notebook
Notebook
Since 2025 · Associate Professor in AI and Actuarial Science
ISFA, Université Claude Bernard Lyon 1
A full data-science track for actuarial students: statistical learning, GLMs, ensembles, interpretability, trustworthy AI and unsupervised methods. Also taught at ENSEA Abidjan.
MSc Actuarial Science Data Science: lectures
Taught in French.
Lecture 1: Apprentissage statistique et lien avec l'actuariat
Cours
Lecture 2: Modèles linéaires généralisés et pénalisés
Cours
Lecture 3: Arbres de décision et méthodes ensemblistes
Cours
Lecture 4: Interprétabilité des modèles d'apprentissage Coming soon
Lecture 5: IA de confiance et biais algorithmiques Coming soon
Lecture 6: Apprentissage non supervisé
Cours
MSc Actuarial Science Actuarial Data Science: practical work
Ten sessions from exploratory analysis to advanced machine learning on real motor insurance portfolios.
TP1: EDA in insurance
Notebook
TP2: Preprocessing
Notebook
TP3: Logistic regression
Notebook
TP4: Actuarial evaluation
Notebook
TP5: GLM versus credibility
Notebook
TP7: Advanced machine learning
Notebook
Mortality score
Notebook
Datasets
freMPLfreMTPL2freqPandas walkthrough
2019 to 2022 · Institut Polytechnique de Paris
Teaching assistant
Tutorials and practical sessions at ENSAE and École polytechnique, alongside the course leads.
1A, semester 1 Algorithme et programmation
Course led by Xavier Dupré.
Feuille de route 2019
Course page
2A, semester 2 Simulation et Monte Carlo
Course led by Nicolas Chopin.
TD1: Loi uniforme et Monte Carlo
R correctionWritten
TD2: Méthode de rejet et Box-Muller amélioré
R correction
TD3: Loi géométrique, variables de contrôle et antithétiques
R correction
TD4: MCMC et importance sampling
R correction
TD5: Cross-entropy et quasi Monte Carlo
R correctionQuasi-MC (Rmd)
2A, semester 2 Theoretical foundations of Machine Learning
3A, semester 1 Advanced Machine Learning
Course led by Vianney Perchet.
PW: Python, linear regression and SVM
Python
PW: RKHS, optimisation and neural networks
WrittenNeural nets
3A, semester 2 Machine Learning for Finance
Course led by Romuald Elie.
Guest speaker on NLP
Milliman × Université Gustave Eiffel
Generative AI hackathon
Crash course & debrief
A fast introduction to retrieval-augmented generation, and the day-one debrief.
Handouts
Cheat sheets
The reference sheets I hand out at the start of every Python course.
Python Jupyter NumPy & SciPy pandas Matplotlib, Seaborn, Bokeh scikit-learn