Sara Bicego
PhD student
Department of Mathematics,
Imperial College London
Learning-based control of multi agent systems
I'm a third year PhD student at Imperial College London, under the supervision of Dante Kalise.
My research interests orbit around scientific computing for high dimensional Hamilton-Jacobi-Bellman equations arising in optimal control. At the core of my PhD project, there are deep learning techniques for taming the curse of dimensionality related to the solution of such PDEs. Current works are focused on mean field control problems for multi agent systems and mean field games.
Short Bio
I'm from a teeny-tiny town in Italy, Recoaro Terme, nestled between green woods and mountains. In 2016 I moved to the city for my B.Sc. in Applied Mathematic and M.Sc in Mathematics at Università degli Studi di Verona.
Now in London, since Autumn 2021, for my PhD (and the sticky toffee puddings).
Conferences and invited talks
Workshop GAMMA24, 24th June 2024, Verona, Italy
New Trends in Optimal Control 15th-17th May 2024, Venice, Italy
Interacting Particle Systems: Analysis, Control, Learning and Computation 6th-10th May 2024, ICERM, Brown University, USA
Junior Analysis Seminar 1st May 2024, ICL, London, UK
Imperial-CNRS Annual Meeting 23rd-24th April, ICL, London, UK
SIAM Conference on Uncertainty Quantification (Data Assimilation and Optimal Control: Theory and Algorithms) 27th February -1st March 2024, Trieste, Italy
International Congress on Industrial and Applied Mathematics (Machine Learning and Differential Equations) 20th-25th August 2023, Waseda University, Tokyo, Japan
Mathematical and Scientific Machine Learning, 5th-9th June 2023, ICERM, USA [poster.pdf ]
British Applied Mathematics Colloquium (Optimisation and control for nonlinear dynamics) 3th-5th April 2023, Bristol, UK
MTNS, September 11th-16th 2022, University of Bayreuth, DE
British Applied Mathematics Colloquium (Mathematical Modelling in the Social Sciences) 11th-13th April 2022, Loughborough, UK
Young Researcher Seminars MAMs, University of Verona, 10th November 2021, Online/Verona, IT
Peer-reviewed publications
Gradient-augmented Supervised Learning of Optimal Feedback Laws Using State-dependent Riccati Equations, G. Albi, S. Bicego and D. Kalise. IEEE Control Systems Letters 6(2022): 836 -841 [arxiv, journal]
Supervised learning for kinetic consensus control, G. Albi, S. Bicego and D. Kalise. Proceedings of the 25th International Symposium on Mathematical Theory of Networks and Systems (I), pp. 308-313 [arxiv, journal]
Preprints
Data/moment-driven approaches for fast predictive control of collective dynamics, G. Albi, S. Bicego, M. Herty, Y. Huang, D. Kalise, C. Segala and D. Kalise [arxiv]
Control of high-dimensional collective dynamics by deep neural feedback laws and kinetic modelling, G. Albi, S. Bicego and D. Kalise [arxiv]
Computation and Control of Unstable Steady States for Mean Field Multiagent Systems, S. Bicego, D. Kalise and G. Pavliotis [arxiv]
