Patent · US Active

Multivariate time-series search and classification learning methods and systems

US12229164B1 · kind B1 · utility

0Cited by
1References
28Claims
0Family size

Assignee

Inventors

Key dates

Filing dateSep 16, 2020
Grant dateFeb 18, 2025
Priority date
Expiry dateSep 16, 2040

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N20/00
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A system for processing multivariate, time-series data to classify portions of the data corresponding to an identified event or condition. Time periods of interest may be identified as search criteria for a variety of purposes. As an example, an event or condition may occur, whereby, precursors to the event or condition may be present in the time-series data, and thereby be a time period of interest as a predictor of future occurrences. As another example, the state of the system during normal operation may be identified. Based on the provided data, the processing system may output a prediction or a likelihood of a re-occurrence of the condition. One or more actions may be taken in response to the output of the system to prevent the occurrence of the condition or otherwise respond to the possibility of the occurrence of the condition.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.