Email:francois.clement@lip6.fr
I’m a third-year PhD Student under the supervision of Carola Doerr in the Operations Research (RO) team in LIP6, Sorbonne Université. This PhD is co-supervised by Luís Paquete at the University of Coimbra. I’m working on discrepancy and more specifically on building low L∞ star discrepancy sets that could be used in practical applications. While our approach is more practical than the usual mathematical perspective on discrepancy, I’m very interested in all the theoretical aspects of the problem: complexity proofs, optimisation formulations and methods and of course developing new algorithms. More generally, I greatly enjoy working on geometric and combinatorial problems!
If you can’t find a paper or need the official version, feel free to send me an email!
2024
Preprint: Outperforming the Best 1D Low-Discrepancy Constructions with a Greedy Algorithm, preprint, arxiv
PhD Thesis: An Optimization Perspective on the Construction of Low-Discrepancy Point Sets, latest version
Preprint: Transforming the Challenge of Constructing Low-Discrepancy Point Sets into a Permutation Selection Problem, preprint, arxiv
2023
Computing Star Discrepancies with Numerical Black-Box Optimization Algorithms, with D. Vermetten, J. de Nobel, A.D. Jesus, C. Doerr and L. Paquete, accepted for GECCO 2023, arxiv.
Preprint: Heuristic Approaches to Obtain Low-Discrepancy Point Sets via Subset Selection, with C. Doerr and L. Paquete, arxiv. Accepted in Journal of Complexity
Preprint: Partitions for Stratified Sampling, with N. Kirk and F. Pausinger, arxiv. Accepted in Monte Carlo Methods and Applications.
Preprint: Constructing Optimal L∞ star discrepancy point sets, with C. Doerr, K. Klamroth and L. Paquete, arxiv
2022:
Star discrepancy subset selection: Problem formulation and efficient approaches for low dimensions, with C. Doerr and L. Paquete, Journal of Complexity, Vol 70, 101645, published, arxiv.
On the two-parameter Erdos-Falconer distance problem in finite fields, with H. Nassajian Mojarrad, D. Hiep Pham and C. Shen, Bulletin of the Australian Mathematical Society, 1-5, published, arxiv.
Distribution of Distances in Five Dimensions and Related Problems with T. Pham, SIAM Journal on Discrete Mathematics, Vol 36, 3, pages 2271-2281, published, arxiv.
06/24: Deyao Chen bachelor student from university of Saint-Andrews, 1 month research internship on algorithms for the L∞ star discrepancy and L2 subset selection.
06/24-12/24: Rabea Freese, master thesis. Co-supervisation with Kathrin Klamroth, University of Wuppertal. Topic: Multi-objective optimization for discrepancies.
A number of the presentations are quite similar so not all the slides are here. If you would like a specific set, reach out by email!
2023
Queen’s University Belfast MSRC Seminar, 06/02/2023, slides
MCM2023, Sorbonne Université, Subset Sampling for Low-Discrepancy point sets and Optimal Constructions, 29/06/2023, slides
GECCO 2023, Lisbon: Computing Star Discrepancies with Numerical Black-Box Optimization Algorithms. Conference presentation linked to the paper above with D. Vermetten, J. de Nobel, A. Jesus, C. Doerr and L. Paquete, slides.
Dagstuhl Seminar 23351: Algorithms and Complexity for Continuous Problems, slides
PGMO23, Optimal Sets for the L∞ Star Discrepancy.
2022
MCQMC Conference in Linz, Austria, Efficient Algorithms for Star Discrepancy Subset Selection, 22/7. slides
PGMO days at EDF Labs in Palaiseau, Subset Sampling for Low-Discrepancy Point Sets, 30/11.
As part of my PhD, I’m teaching undergraduate courses in TD and TME in Sorbonne Université. TD correspond to exercice sessions on paper and TME to exercice sessions/projects on computer. French Bachelors are in 3 years and usually followed by 2 years of Master.
2022-2023
LU3IN003 Algorithmic 2 (Bachelor third year)- TD: General programming paradigms such as dynamic programming, greedy algorithms and common graph algorithms (Kruskal, Prim, Dijkstra).
LU2IN005 Discrete Mathematics (Bachelor second year) -TME: A course on common mathematical reasoning in computer science: logic, induction and a large part on automata. I was in charge of project supervision for one group: students had to code common functions on automata.
LU2IN003 Algorithmic 1 (Bachelor second year) -TD: A basic introductory class to algorithms, with basic complexity and validity proofs as well as the discovery of graphs.
2021-2022
LU2IN005 Discrete Mathematics (Bachelor second year) -TME: See above but with two groups
LU1IN001 Introduction to programming (Bachelor first year) - TME: An introductory class to programming in Python.
LU2IN003 Algorithmic 1-TD: See above.
Reviews for Journal of Complexity, Optimization and Engineering, IEEE Transactions on Evolutionary Computation, Theory of Probability and Mathematical Statistics
Organized the LIP6 RO team seminar in 2024, with a presentation every week from February to July.
I did a classe préparatoire in Lycée Hoche in Versailles (MPSI-MP*), followed by École Polytechnique where I obtained a “diplôme d’Ingénieur de l’École Polytechnique”, with a focus on theoretical computer science and optimisation. For those not familiar with the French system, classe préparatoire+ engineering school is roughly ther same as a Master’s degree, but with a bigger range of topics. I also studied mathematics, physics, applied maths and a number of history and politics classes in École Polytechnique.
As a complement and as part of my last year in Polytechnique, I obtained a Master in Computer Science from ETH Zürich in 2022, with a specialisation in Theoretical Computer Science. I did my Master thesis with Dr. Thang Pham (now at VNU in Hanoi) on the Erdos-Falconer distance problem in finite fields.
A relatively up-to-date CV, the above list is more accurate for publications.