CV

Damien Aupretre

Student

📞 +33 7 67 91 90 60
✉️ damien.aupretre@student-cs.fr
🐙 TheTrigun99 - github

Education

Master of Engineering — CentraleSupélec (2024 – 2027)

Part of Université Paris-Saclay, ranked #12 worldwide (QS 2024).
Relevant coursework: measure theory & integration, PDEs, information theory, fluid mechanics, signal processing, quantum physics, statistics, Stochastic processes.

CPGE PC* — Lycée Kléber (2022 – 2024)

Two-year intensive program in mathematics and physics, preparing for France's most selective engineering schools. Admitted to École Polytechnique (ranked #1 in France, top 0.1% of applicants), CentraleSupélec, Mines PSL (Most prestigious engineering schools)
GPA: 4.3 / 4.3

Deutsch-Französisch Gymnasium (DFG-LFA) (2015 – 2022)

Bilingual French-German Abitur : 1.0 (maximum grade)
Specialisation: mathematics, biology, and chemistry

Projects-Experience

Explaining and implementing Research Papers in Machine learning — Personal Project (2025 - ??)

Focusing on a deep understand of machine learning algorithms/optimization and implementing these from scratch
Implemented Word2Vec (SGNS) from scratch, including the skip-gram objective and negative sampling.
Implemented the Safe Screening Gap Rule for LASSO — an optimisation technique that provably eliminates inactive features before convergence, dramatically reducing computation.
Built core regularisation methods (LASSO, Ridge)

Kaggle Challenges — Personal Project (2025)

Learned the machine learning basics and Deep Learning basic library and participated in 2 competitions
Predicted accident risk from tabular data using an ensemble of XGBoost, CatBoost, and neural networks.
Built a full ML pipeline covering feature engineering, hyperparameter tuning, and model stacking.

Health Data Science Project — CentraleSupélec (2025)

Analysed high-dimensional multimodal genomic data (gene expression + DNA alteration) from pediatric brain cancer patients to identify tumour-localisation biomarkers.
Implemented and compared multiple sparse classification methods (LASSO logistic regression, Sparse Discriminant Analysis, SGCCA, Multiview), with a focus on the multi-block methods handling joint analysis across data types.
Reduced dimensionality from 15,000+ genomic variables down to a handful of robust, interpretable biomarkers per tumour class.

TIPE, Independent Physics Research Project — Lycée Kléber (2023)

Investigated capillarity and buoyancy effects by varying liquid volumes and surface properties of ping-pong balls.
Modelled ejection dynamics from a water-filled cup using fluid mechanics principles (surface tension, pressure balance).

Skills

Skills

Python (NumPy, pandas)
Machine learning
Linear Algebra
R (statistical computing)
Stochastic processes
Communication

Languages

Languagues

French — Native
German — Fluent (C1-certified)
English — Fluent