The scope of the symposium covers all major aspects of system identification, machine learning techniques for decision and control of dynamical processes, experimental modelling, signal processing, adaptive control and reinforcement learning. Theoretical, methodological and scientific developments involve a large variety of application areas. To enhance the applications and industrial perspective of the symposium, participation by authors from industry is particularly encouraged.

Relevant topics for the symposium program include:

  • Identification of linear, nonlinear, time varying, multivariable, hybrid and distributed systems

  • Black-box modeling (neural and deep networks, kernel-based approaches, support vector machines, kriging ... )

  • Machine learning, data mining and Bayesian approaches

  • Linear and nonlinear time series analysis

  • Estimation from spatio-temporal data

  • State estimation and parameter tracking

  • Robustness issues in identification

  • Sequential Monte Carlo methods, including particle filtering

  • Parameter estimation and inverse problems

  • Modeling and identification of quantized systems

  • Identification for control, adaptive control, data-based controller tuning, reinforcement learning

  • Statistical analysis and uncertainty characterization

  • ldentifiability

  • Experiment design

  • Model validation

  • Sparse estimation and other types of regularization

  • Identification and estimation in data rich (big data) and networked envi- ronments

  • Monitoring and fault detection

  • Applications (including but not limited to transportation, telecommunications, aerospace, automotive, process control, motion control, robotics, econometrics, modal analysis, bioengineering and medical systems, ecosystems, energy and information networks)

  • Teaching  identification.