Name
Abstract | ||
|---|---|---|
Gaussian process model provides a flexible, probabilistic, non-parametric model. Many examples for system identification using Gaussian process have been reported and verified. However, for real plants that have secular change or whose properties change upon time, it is difficult to apply standard Gaussian process and it is important to keep the uncertainties of the modeling properly. In this paper, we consider a system identification for nonlinear time-varying systems using recursive Gaussian process (RGP). We propose two methods for this problem. One is RGP for long-term prediction, and the another is robust RGP for outliers. The effectiveness of the proposed methods will be shown by numerical simulations. |
Year | Venue | Field |
|---|---|---|
2017 | 2017 AMERICAN CONTROL CONFERENCE (ACC) | Nonlinear system,Control theory,Computer science,Robustness (computer science),Gaussian process,Artificial intelligence,Probabilistic logic,System identification,Recursion,Numerical models,Algorithm,Outlier,Machine learning |
DocType | ISSN | Citations |
Conference | 0743-1619 | 0 |
PageRank | References | Authors |
0.34 | 6 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kwangwoo Seo | 1 | 0 | 0.34 |
| Masaki Yamakita | 2 | 266 | 57.24 |