Program

PROGRAM OF THE 5th SPRING SCHOOL on

Data-Driven Model Learning of Dynamic Systems

UPDATE 11 March 2022: The program and the zoom link to attend the lectures have been sent to the participants on Friday 11 March 2022. If you did not receive this mail, please, contact the organizers.

Basics of linear system identification 

Lectures on Monday 4 April (afternoon) and on Tuesday 5 April (afternoon)

Exercises on Wednesday 6 April (morning)

Lecturer: Xavier Bombois, CNRS Research Director, Laboratoire Ampère, Ecole Centrale de Lyon 

Theme 1: Introduction;concepts; identification cycle

Theme 2: Parametric (prediction error) identification methods: prediction criterion and model structures, linear and pseudo-linear regressions, conditions on data, statistical and asymptotic properties, model set selection and model validation

Theme 3: Non-parametric identification (ETFE)

Theme 4: Experiment design.

Closed-loop identification and Design of optimal identification experiments 

Lectures on Wednesday 6 April (afternoon)

Exercises on Thursday 7 April (morning)

Lecturer: Xavier Bombois, CNRS Research Director, Laboratoire Ampère, Ecole Centrale de Lyon

Theme 1: Closed-loop identification: different methods, informativity, ...

Theme 2: Optimal experiment design: formulation as an optimization problem, accuracy and cost constraint

Theme 3: Optimal experiment design: convexification of the optimization problem, parametrization of the to-be-design power spectrum

Theme 4: Optimal experiment design: Alternative formulations, least costly experiment design

Dynamic model learning

Lectures on Thursday 7 April (afternoon) and on Friday 8 April (afternoon)

Lecturer: Håkan Hjalmarsson, Professor, KTH, Stockholm, Sweden

Theme 1: Fundamental parameter estimation concepts: Sufficient statistics, the Cramér-Rao bound, the maximum likelihood estimator, estimator-based methods

Theme 2: Minimum MSE estimators: Bayes estimators, empirical Bayes methods, risk estimation methods, Gaussian processes, asymptotic analysis

Theme 3: Application to dynamical models: linear models, non-linear input-output models, non-linear state-space models

Theme 4: Computational tools: Sampling, Markov Chain Monte Carlo methods, particle filtering and smoothing