| Abstract: |
| Singular spectrum analysis (SSA) as a data-adaptive multimodal method based on principal component analysis (PCA) decomposes the lagged copies of process signals into different components such as trend and noise using their covariance structure. SSA has demonstrated its capability in chemical and metallurgical process monitoring using multivariate statistics. I will be presenting the work on the recent applications of variants of SSA in nonlinear process monitoring. |
|