|
Abstract:
The goal of this research effort is to investigate methods to fuse vast amounts of data coming from different sensor sources with a multi-layered semi-supervised learning approach. This approach will use basic statistical techniques to identify key predictors, some correlation techniques to validate the source, quality and temporal aspects of the data, artificial neural networks for troubleshooting sources of system variability, and semi-supervised learning techniques which will provide adjustable thresholds for forecasting and detecting various anomalies or events of interest.
| Limitations: |
 APPROVED FOR PUBLIC RELEASE |
| Description: |
Final technical rept. Jul 2009-Dec 2010 |
| Pages: |
100 |
| Report Date: |
JUN 2011 |
| Report Number: |
A141645 |
|
|
|
|
|
Keywords relating to this report:
Email This Abstract
