Methods involving computing correlation anomaly scores
US7483934B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 18, 2007 |
| Grant date | Jan 27, 2009 |
| Priority date | — |
| Expiry date | Dec 18, 2027 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06V10/95
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
An exemplary method for computing correlation anomaly scores, including, defining a first similarity matrix for a target run of data, the target run of data includes an N number of sensors, defining a second similarity matrix for a reference run of data, the target run of data includes the N number of sensors, developing a k-neighborhood graph Ni of the i-th node for the target run of data, wherein the k-neighborhood graph of the i-th node is defined as a graph comprising the i-th node and its k-nearest neighbors (NN), developing a k-neighborhood graph Ni of the i-th node for the reference run of data, defining a probability distribution p(j|i), wherein p(j|i) is the probability that the j-th node becomes one of the k-NN of the i-th node, coupling the probability between the i-th node and the neighbors of the i-th node, determining an anomaly score of the i-th node, and determining whether the target run of data has changed from the reference run of data responsive to determining the anomaly score of the i-th node.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.