Abstract
The problem of determining information-bearing sensors in the presence of multiple field sources and (non-)linear data models is considered. To this end, a novel canonical correlation analysis (CCA) framework combined with norm-one regularization is introduced to identify correlated measurements across the distributed sensors and cluster the sensor data based on their source content. A distributed algorithm is also put forth for informative sensor identification in nonlinear settings using the novel CCA approach. Toward this end, the sparsity-aware CCA framework is reformulated as a separable constrained minimization problem which is solved by utilizing block coordinate descent techniques combined with the alternating direction method of multipliers. Numerical tests demonstrate that the distributed sparse CCA scheme put forth outperforms existing alternatives when it comes to clustering the sensor data based on their source content.
Original language | English (US) |
---|---|
Title of host publication | Conference Record of the 47th Asilomar Conference on Signals, Systems and Computers |
Publisher | IEEE Computer Society |
Pages | 639-643 |
Number of pages | 5 |
ISBN (Print) | 9781479923908 |
DOIs | |
State | Published - Jan 1 2013 |
Externally published | Yes |
Event | 2013 47th Asilomar Conference on Signals, Systems and Computers - Pacific Grove, CA, United States Duration: Nov 3 2013 → Nov 6 2013 |
Other
Other | 2013 47th Asilomar Conference on Signals, Systems and Computers |
---|---|
Country/Territory | United States |
City | Pacific Grove, CA |
Period | 11/3/13 → 11/6/13 |
Keywords
- canonical correlation analysis
- Distributed processing
- sparsity