My main research is on numerical optimization: theory, algorithms, and applications. I am currently focusing on efficient methods (both deterministic and stochastic), with applications in machine learning and data science.
Currently I am a postdoctoral researcher at university of Helsinki, Finland, working in the department of Mathematics and Statistics with Tuomo Valkonen.
Area of intrerest:
I am broadly interested in exploring the following areas:
Optimization
- Efficient methods for (non)convex problems, (non)smooth formulations
- Structure-exploiting methods (sparsity, stochasticity)
- Second and higher-order optimization methods
- Online optimization
Machine Learning
- Applications of optimization in machine learning
- Distributed and Federated learning
- Communication- and computation-efficient algorithms
I defended my Ph.D thesis in May 2024, supervised by Ion Necoara, within the Marie Skłodowska-Curie Actions (MSCA) as an Early Stage Researcher, TraDE-OPT H2020 ITN project, as ESR 10. My thesis “Higher-order methods for composite optimization and applications”, focuses on advanced optimization techniques and their applications.
I received a BSc degree in Applied Mathematics from University Cadi Ayyad in 2016. I obtained a MSc degree in Applied mathematics from Faculty of Science and Technology, Morocco in 2018.
Publications:
2025
- Y. Nabou, Lahcen El Bourkhisi, Sebastian U. Stich, Tuomo Valkonen, Monotone and nonmonotone linearized block coordinate descent methods for nonsmooth composite optimization problems, arXiv.
- Y. Nabou, Nonmonotone higher-order Taylor approximation methods for composite problems, arXiv.
2024
- Y. Nabou, Ion Necoara, Regularized higher-order Taylor approximation methods for composite nonlinear least-squares, arXiv.
- Y. Nabou, Ion Necoara, Moving higher-order Taylor approximations method for smooth constrained minimization problems, SIAM Journal on Optimization (accepted), arXiv.
- Y. Nabou, Francois Glineur, Ion Necoara, Proximal gradient methods with inexact oracle of degree q for composite optimization, Optimization Letters, (pdf).
- Y. Nabou, Ion Necoara, Efficiency of higher-order algorithms for minimizing composite functions, Computational Optimization and Applications, (pdf).
- Y. Nabou, Lucian Toma, Ion Necoara, Modified projected Gauss-Newton method for constrained nonlinear least-squares: application to power flow analysis, European Control Conference, (pdf).
Talks & Conferences:
- June 29 - July 2, 2025: Nonmonotone linearized block coordinate descent methods for nonsmooth composite problems, EUROPT 2025, Southampton, UK.
- October 31, 2024: Higher-Order Methods for Composite Optimization with Application, MOP Research Seminar, Saarland University, Germany (online).
- September 16, 2024: Efficient Algorithms for Composite Problems with Applications, ESAT KU Leuven, Belgium, invited by Hakan Ergun (online).
- January 23-26, 2024: Moving Higher-Order Taylor Approximations Method for Smooth Constrained Minimization Problems, Workshop on Analysis and Potential, event website, Bucharest, Romania.
- September 29-30, 2023: Moving Higher-Order Taylor Approximations Method for Smooth Constrained Minimization Problems, Conference on Statistical Modeling with Applications, event website, Bucharest, Romania.
- June 13-16, 2023: Modified Projected Gauss-Newton Method for Constrained Nonlinear Least-Squares: Application to Power Flow Analysis, European Control Conference (ECC23), conference website, Bucharest, Romania.
- July 4-8, 2022: Efficient Optimization Methods for Complex Systems, Workshop on Algorithmic and Continuous Optimization, event website, UCLouvain, Belgium.
- August 24, 2021: Higher-Order Algorithms for Composite Minimization Problems, MaLGa Machine Learning Genoa Center, Italy, invited by Silvia Villa (online).
Teaching experience:
- Spring 2026 (upcoming) – Appointed to teach MAT11015: Basics of Mathematics in Machine Learning II, University of Helsinki.
- Spring 2025 – Course Assistant for MAT11015: Basics of Mathematics in Machine Learning II, University of Helsinki.