Computation-enabled design, optimization, and data science

Description

This research area deals with methodological and applicative problems in optimization,  data analysis and optimal design. In light of the present availability of large amounts of data and the opportunities offered by Artificial Intelligence, an important part of the investigation is carried out in a data-driven fashion. Other relevant methodologies reside in convex optimization, scenario approach and probabilistic reasoning, predictive models. Typical applications are in the fields of financial engineering, optimization of complex systems (such as mobility systems, electrical vehicles and fleets), human-vehicle interaction, logistics, decision support systems, urban systems.

A particular promising research field in this setting involves application of Artificial Intelligence, and especially its machine learning embodiment for automated design of circuits and EM devices, such as metasurface antennas. Notable examples for this task are the so-called Physics-informed design algorithms, which exploit fast forward solvers combined with sophisticated optimization tools to provide full design and layout in a very reduced runtime.

ERC sectors 

  • PE7_1 Control engineering
  • PE7_3 Simulation engineering and modelling
  • PE7_6 Communication systems, wireless technology, high-frequency technology
  • PE1_20 Control theory, optimisation and operational research
  • PE6_7 Artificial intelligence, intelligent systems, natural language processing
  • PE6_12 Scientific computing, simulation and modelling tools

Keywords 

  • Optimization
  • Data science
  • Data analytics
  • Machine learning
  • Probabilistic control
  • Automated design
  • Inverse design