Learning complex systems from data: the set membership approach
The paper deals with the problem of making inferences on unknown nonlinear systems based on finite and noise-corrupted measurements. The approach is based on a set membership framework and does not need assumptions on the functional form of the regression function describing the system. Necessary and sufficient conditions for checking the assumptions' validity are given and almost optimal algorithms are presented for the cases where the desired inferences are identification and prediction. For the entire collection see [Zbl 1012.00049].
- Unified set membership theory for identification, prediction and filtering of nonlinear systems
- Set membership identification of nonlinear systems
- scientific article; zbMATH DE number 1057954
- Set-membership identification of systems with parametric and nonparametric uncertainty
- scientific article; zbMATH DE number 1552745
- On the inference of semi-coherent structures from data
- Nonlinear set membership prediction of river flow
- Model-free approaches in learning the multivariate linear regressive models
- On system identification of complex systems from finite data
- scientific article; zbMATH DE number 926776 (Why is no real title available?)
- Using statistical learning theory to rationalize system model identification and validation. I: Mathematical foundations
- Unified set membership theory for identification, prediction and filtering of nonlinear systems
- Data mining and knowledge discovery via statistical mechanics in nonlinear stochastic systems
This page was built for publication: Learning complex systems from data: the set membership approach
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4452934)
