MSc student in Quantitative Economics (Data track), with a BSc double degree in Economics & Mathematics (Université Paris-Saclay). I like problems where statistics, probability and code meet: building estimators from scratch, simulating stochastic systems, and checking results against theory. Open to internships in data science, machine learning and quantitative research.
| Project | What it shows | Stack |
|---|---|---|
| Exact mixing times of card shuffles | Markov chains on S₅₂: exact total-variation and separation curves via a lumping argument (no enumeration of 52! states), sharp vs. textbook strong stationary times, Bayer–Diaconis riffle formula in exact arithmetic, cut-off phenomenon | Python · NumPy · pytest |
| Abelian sandpile on ℤ² | Proofs of the abelian property and propagation bounds, verified by a vectorised simulator; exact Green-function identities; self-organised criticality and power-law tail estimation (MLE) | Python · NumPy · SciPy |
| Asylum flows: gravity model with HDFE | Panel econometrics on 39k observations: from-scratch FE-OLS / FE-PPML with clustered SEs, replication of fixest, detection of a coding error and a functional-form artifact (log(1+y) vs. PPML) |
Python · R (fixest) |
Python (NumPy, pandas, SciPy, Matplotlib, pytest) · R (tidyverse, fixest) · LaTeX · Git & GitHub Actions
Methods: statistical modeling & inference (maximum likelihood, GLMs, regression with high-dimensional fixed effects, cluster-robust standard errors) · probability & stochastic processes (Markov chains, Monte Carlo simulation, heavy-tailed distributions) · reproducible research (unit tests, CI, documented pipelines).