Abstract: |
The wide deployment and application of distributed sensing and computer systems have resulted in multi-stream sensing data, which provides unprecedented opportunities for understanding and improving operations of complex systems. However, it also brings out new research challenges on data analysis due to high-dimensional and complex temporal-spatial correlated data structure. In this talk, as an example, I will discuss a critical research issue on how to separate immeasurable embedded individual source signals from the mixed sensor measurements. In this research, a new method is proposed by integrating Independent Component Analysis (ICA) and Sparse Component Analysis (SCA) for source signal separations. Going beyond the existing ICA method, the proposed method can estimate not only independent source signals but also the dependent source signals if they have dominant components in either the time or linear transform domains. Based on those identified source signals, it can facilitate directly monitoring of individual source signals with explicit diagnostic information. |