Efficient sampling with replacement
US9135300B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 20, 2012 |
| Grant date | Sep 15, 2015 |
| Priority date | — |
| Expiry date | Apr 4, 2034 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06F16/2462
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A method, article of manufacture, and apparatus for process data. In some embodiments, this includes determining a dataset, wherein the dataset includes a plurality of data values and a plurality of identifiers, and wherein the plurality of identifiers are in a non-continuous order, and wherein the dataset is stored across a number of segments, determining a number of samples to take from the dataset, identifying a minimum identifier and a maximum identifier, and a total number of identifiers, calculating a density based on the plurality of identifiers, minimum identifier and maximum identifier, calculating a number of samples to take from a segment based on the calculated density, the number of samples to take from the dataset, and the number of segments, generating a sub-dataset based on the calculated number of samples to take from a segment, joining the generated sub-dataset with the dataset, and storing the joined dataset in a storage device.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.