Dante Kalise

Senior Lecturer in Computational Optimisation and Control

Department of Mathematics, Imperial College London

Control, Optimization, and Scientific Computation:

My research revolves around scientific computation methods for optimization and control, nonlinear PDEs, high-dimensional approximation and agent-based models across scales. Some relevant keywords below.

Foundations: Dynamic optimization with ODE/PDE constraints; large-scale and sparse optimization; mathematical control theory; high-dimensional approximation; numerical methods for Hamilton-Jacobi PDEs and dynamic programming; optimal transport, mean-field control/games.

Applications. Scientific Machine Learning: data-driven control design in trajectory optimization, epidemiology, and large-scale dynamics. Agent-based models: nonlocal PDES, collective behaviour, consensus control. Nonlinear control design: power electronics, swarm robotics, reinforcement learning. Control of PDEs: fluid flow and vibration control, optimal actuator design.

Please have a look at Research , Publications and CV for more details. I'm always happy to discuss about any of these topics for projects and supervision, so please feel free to get in touch!

Short Bio

Since July 2021 I'm Senior Lecturer in Computational Optimisation and Control at Department of Mathematics, Imperial College London. My criminal record includes: Assistant Professor in Applied Mathematics, University of Nottingham (2019-2021), Imperial College Research Fellow (Maths, 2017-2019), Research Scientist (Optimization and Control Group, 2013-2017), RICAM Linz, Austrian Academy of Sciences, and Postdoctoral Researcher (Maths, EU ITN-SADCO, 2011-2013), Sapienza University of Rome.

I'm originally from Chile and grew up in front of the Pacific, in Viña del Mar. I received a B.Sc. in Mathematical Engineering and a M.Sc. in Electronic Engineering from the Federico Santa María Technical University in Valparaíso. In 2008 I moved to Norway, where I got a Ph.D. in Mathematics in 2012 from the University of Bergen (with some periods away from the rain in Oslo and Trento).


Past News and Miscellanea

  • Together with Timm Faulwasser and Emmanuel Trélat we are organizing two MCSS special issues on "Machine Learning for Control Systems and Optimal Control” and “Optimal Control and Dynamic Games: Large Time Behavior and Geometry”. If you are interested in contributing, please get in touch.

  • Our MTNS short course on Modelling, Simulation, and Control of Social Dynamics Across Scales with Greg Pavliotis and our session on Optimal Actuator/Sensor Design with Kirsten Morris have been accepted but postponed to 2021.

  • I joined the Editorial Board of Mathematics of Control, Signals, and Systems (Springer) , submissions on optimal control are particularly welcome!

  • Together with Karl Kunisch we organized a workshop on Feedback Control as part of the Special Semester on Optimization at RICAM.

  • Together with Giacomo Albi, Michael Herty and Giuseppe Visconti we organized a 4-session minisymposium at ICIAM 2019 on Novel Concepts in Model-driven Optimization and Control of Agent-based Systems.

  • Together with Mattia Zanella we organized a minisymposium at the 28th Biennial Numerical Analysis Conference on Computational methods for model-driven optimization and control under uncertainty.

  • The book on Numerical Methods for Hamilton-Jacobi-Bellman equations is finally available!

  • Starting from October 12th 2018, I'll be teaching a TCC course at Imperial on Mathematical Foundations of Reinforcement Learning, please don't hesitate to get in touch if you are interested in attending in person or holographically!.

  • Since April 2018 I am also affiliated to the Data Science Institute at Imperial.

  • The book on Optimal Control and Applications is finally out!.

  • An article by the Austrian Press Agency about our research on computational approaches to the control of social dynamics and collective behavior is available here.

  • The Workshop on Numerical methods for Hamilton-Jacobi equations in optimal control and related fields, November 21th-25th 2016 at RICAM was a great success. We had more than 30 international speakers who presented the state of the art concerning the numerical approximation of HJB eqns. and related topics. Thanks to all of them for their participation!

  • You can find here a talk that I gave in August in Jyväskylä about our current research on high-dimensional HJB equations. I wish I would have known the camera was recording the center of the classroom.

  • Emiliano Cristiani did a press release about our paper "Invisible control of self-organizing agents leaving unknown environments", which is now available in several Italian media.

  • I'll just leave this here.

Past conferences and talks