A Seminar by prof. Roberto Diversi. Curriculum: Automatic Control and Operational Research.
Date:
Event location: Teams
Type: Course
The problem of deriving possible linear relations from data affected by additive noise will be considered, with reference to both the static and the dynamic case.
Traditional approaches like least squares lead to unique solutions because of the assumptions ('prejudices') behind the noise corrupting the data. A more general context is that of the errors-in-variables models where all the involved variables are assumed as affected by noise.
The seminar will be focused on errors- in-variables estimation schemes. In particular, the Frisch scheme, leading to a whole family of models compatible with a set of noisy data, will be considered.
The seminar will be given through Teams.
In order to be added to the Seminar Team, please send an email to prof. Roberto Diversi.