Comparing data series associated with two systems to identify hidden similarities between them
US8738563B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Mar 28, 2010 |
| Grant date | May 27, 2014 |
| Priority date | — |
| Expiry date | Apr 12, 2031 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N7/01
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A method that includes: sampling a plurality of parameters associated with a first and a second system to yield a first and a second set of data series, respectively, each data series being associated with samples of a respective parameter; applying, for each two respective data series, a transformation associated with a specified order selected from a plurality of ordered transformations, such that the selected transformation is applied repeatedly to at least one of the two respective data series, wherein, in each repeated application, the transformation is assigned with a different value of a transformation parameter; determining, for each parameter, a one of the plurality of transformation parameters for which the correlation level of the two respective time series is above a specified value; and repeating the applying and the determining with a transformation that is associated with a higher order of the ordered transformations.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.