Full name: Alexandre Magueresse
Date of birth: April 21, 1998 in Pontivy, France
Nationality: French
Institutional e-mail: alexandre.magueresse@monash.edu
Github: AlexandreMagueresse
ORCID: 0000-0002-6296-5399
Functional analysis, Partial differential equations, Ordinary differential equations.
Finite Elements, Virtual Elements, Isogeometric Analysis, Numerical integration.
Tensor-product methods, Sparse grids.
Nonlinear approximation, Adaptive methods, Mesh adaptivity, Scientific Machine Learning.
Open source numerical software, Julia.
PhD in Mathematics. Feb. 2022 – Aug. 2025, Monash University (Melbourne, Australia).
PhD thesis: Energy minimisation in nonlinear spaces for the approximation of general PDEs. Under the supervision of Prof. Santiago Badia, Prof. Tiangang Cui, Dr. Jai Tushar.
This thesis investigates nonlinear methods for approximating solutions to elliptic partial differential equations via energy minimisation. The first part introduces adaptive quadratures for physics-informed neural networks, leveraging piecewise-linear approximations of activation functions to enable accurate integration with few evaluation points. The second part presents a general optimisation framework for nonlinear approximation spaces, establishing convergence under broad structural conditions. The framework is realised using free-knot B-splines, where knot positions adapt to the local features of the solution. Numerical experiments reveal marked improvements in accuracy compared to conventional linear methods, highlighting the potential and challenges of nonlinear discretisations in PDE solvers.
Relevant courses: Differential Geometry (Prof. Todd Oliynyk); Adaptive Finite Elements (Prof. Carsten Carstensen); Fixed-point arguments for coupled PDEs (Dr. Segundo Villa-Fuentes), Malliavin calculus (Dr. Kaustav Das).
Master in Advanced Computing, specialisation in Artificial intelligence. Aug. 2019 – Jul. 2021, Tsinghua University (Beijing, China). GPA : 4.0, Ranking : 1/141.
Master thesis: DepBERT: Towards a Grammatical Dependency Enhanced BERT. Under the supervision of Prof. Maosong Sun.
Addition of a new training task for the BERT model to enrich its linguistic knowledge; Modified BERT's masking strategy to give more weight to words involved in subject-verb or verb-object relationships, particularly when the two words are far apart; Use of the Penn Tree Bank database, and creation of a new database using an automatic parser on the Wikipedia corpus; Improved ability to reproduce a syntax tree by 1% (8.3% error reduction) while maintaining accuracy on the GLUE benchmark.
Relevant courses: Computational linguistics; Natural Language Processing; Combinatorics and algorithm design; Time series analysis; Machine Learning; Deep Learning; Web Information Retrieval; Introduction to Big Data systems.
Master of Engineering. Sep. 2017 – Jul. 2021, École Centrale de Lyon (Lyon, France). GPA: 3.9, Ranking : 10/400.
Relevant courses: Numerical analysis; Finite Element Method; Algorithm design; Data structures; Collaborative algorithms; Natural language processing; Image processing; Signal processing; Continuum mechanics; Fluid mechanics.
Classes préparatoires aux grandes écoles (CPGE), MPSI – MP*. Sep. 2015 – Jul. 2017, Lycée Chateaubriand (Rennes, France).
Two-year intensive program in mathematics, physics, chemistry and computer science (Python and OCaml) to prepare for the entrance exam to French engineering schools.
Postdoctoral researcher Apr. 2026 – Present, Institut für Geometrie und Praktische Mathematik (IGPM), RWTH Aachen University (Aachen, Germany)
Approximation of high-dimensional PDEs by low-rank tensors and neural networks, under the supervision of Prof. Markus Bachmayr.
Research assistant. Aug. 2025 – Oct. 2025, Monash University (Melbourne, Australia).
Research on nonlinear optimisation under the supervision of Dr. Janosch Rieger.
Middle school mathematics teacher. Sep. 2021 – Jan. 2022, École Notre-Dame de la Providence (Vincennes, France).
Substitution in mathematics for grades 6 to 9; Substitution in digital sciences and technology for grade 10; Teaching in 7 classes simultaneously, one of which as the form teacher.
Research engineer in Deep Learning. Mar. 2021 – Jan. 2022, EU H2020 NeurONN (Paris, France). NeurONN is a research project funded by H2020 EU's research and innovation programme with core subject "Energy-efficient bio-inspired devices accelerate route to brain-like computing".
Implementation of a digital version of the oscillatory neural networks algorithm (Hopfield networks) in Tensorflow (Python); Performance evaluation and comparison on several tasks: digits recognition (MNIST), sound localisation; Evaluation of the robustness to noise; Development of a hierarchical variant aiming to cluster reference patterns.
Deep learning intern. Jan. 2021 – Jul. 2021, A.I. Mergence (Paris, France).
Implementation of a Deep Learning algorithm for environmental sound classification and recognition (Python, TensorFlow); Collaboration on a European project called NeurONN, designing a new neural network architecture (Python, TensorFlow); Involvement in the RoboVox project : robust vocal identification for mobile security robots, funded by the French Research Agency (ANR), creation of a voice corpus (Python, SQL, React JS); Adaptation for a communication protocol between several languages in order to share code across platforms (Python, JS, C++); Optimisation of a simulation algorithm for the autonomous exploration of a space to be protected (Unity, C#); Renovation of the company's website (React JS).
English class teaching assistant. Sep. 2020 – Jan. 2021, Tsinghua University (Beijing, China).
Mentoring of groups of bachelor and master students to develop their oral skills in English (speaking and listening).
Machine Learning intern. May. 2019 – Jul. 2019, Cybel Angel (Paris, France).
Development and implementation of machine learning algorithms for classification and clustering; Comparison and testing of models in a production environment, measurement of their performances.
Postgraduate student representative for the School of Mathematics at Monash University (2023, 2024, 2025)
Reviewer for Computers & Mathematics with Applications (2023, 2024; 4 reviews)
Mathematics Learning Centre. Monash University, Semester 2 2024.
Foundation mathematics (ENG 1090). Monash University, Semester 1 2024.
Numerical linear algebra (MTH 3320). Monash University, Semester 1 2024 (only marking).
Engineering mathematics (ENG 1005). Monash University, Semester 2 2023.
Research program on the Numerical Analysis of Interface and Multiphysics Problems, May 2025, University of Melbourne, Creswick campus, Australia.
ECCOMAS Artificial Intelligence and Computational Methods in Applied Science (AICOMAS), Feb 2025, Arts et Métiers Institute of Technology, Paris, France.
Computational Techniques and Applications Conference (CTAC), Nov 2024, Monash University, Melbourne, Australia.
VicANZIAM One Day Meeting, June 2024, RMIT University, Melbourne, Australia.
Workshop on Numerical Analysis of Partial Differential Equations (WONAPDE), Jan 2024, Universidad de Concepción, Concepción, Chile.
Literature review for the automatic processing of low-resource natural languages, Tsinghua University, 2020. See our preprint Low-resource Languages: A Review of Past Work and Future Challenges. Constitution and review of the state of the art by subtask (parts of speech tagging, dependency parsing, named entities recognition, tagging and linking, speech recognition, translation, classification, embedding) of NLP methods used in the context of low-resource languages; Proposals for new resources and research directions.
Automatic image colouring, Tsinghua University, 2020. Coloured black and white image based on context and user-specified cues. Modification of an existing algorithm by Zhang & al. (Real-Time User-Guided Image Colorization with Learned Deep Priors, 2017): instead of prescribing colours at points, the user prescribes colours on lines; Improved the signal to noise ratio and produced more realistic images.
Artificial interpretation of natural languages, École Centrale de Lyon, 2019. Design and implementation of an automatic summarisation algorithm from English corpora (Python); Synthesis of the English grammar (according to Chomsky's formalism) in order to implement an English parser for a hierarchical combination of word vectors into a sentence vector.
Graphic interface and artificial intelligence for the Penguins game, École Centrale de Lyon, 2018. The Penguins game is a board game for 2 to 4 players, where the objective is to recover as many points as possible by eating fish on a 60-square board. Each move removes one square, so you have to reconcile mobility with the lure of winning. Development of a digital version of the game; Implementation of several artificial intelligences at different levels, based on tree exploration principles: minimax, maxn algorithm, alpha-beta pruning, Monte-Carlo tree search (MCTS) guided by a random forest, heuristics specific to the Penguins game. Six-person teamwork, Python, Pygame, GitHub.
Reconstruction of the score of a piano recording, Lycée Châteaubriand, 2017. Design of a strategy to separate note groups (temporal and amplitude conditions); Temporal filtering with various windows; Fourier analysis of a sound recording followed by harmonic analysis to identify the notes played; Reconstruction of the score.
Native: French.
Advanced: English (C2, IELTS 8.0), Spanish (C2), Chinese (C1, HSK 5).
Intermediate: German (B1), Russian (B1), Esperanto (B1).
Scientific computing: Julia, Python, Matlab, some experience with C++, Fortran, OpenMP.
Machine learning: Flux, Knet (Julia); Pytorch, Tensorflow, Scikit-Learn (Python).
Web Development: HTML5, CSS3, JavaScript, NodeJS, VueJS, MySQL.