Patent · US Active

Information processing apparatus, and method

US10795326B2 · kind B2 · utility

0Cited by
3References
12Claims
0Family size

Assignee

Inventors

Key dates

Filing dateNov 24, 2017
Grant dateOct 6, 2020
Priority date
Expiry dateNov 24, 2037

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N99/00
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

The present disclosure relates to an information processing apparatus, a method, and a program capable of causing a system to efficiently learn a method of controlling a person. A control learning system calculates a reward based on an input objective state of a control target and a state of the control target based on a sensing result of the control target. The control learning system performs reinforcement learning using the calculated reward and the state of the control target to select a better action for bringing the control target closer to the objective state. The control learning system executes the selected action for the control target. For example, the present disclosure can be applied to a control learning system including a terminal and a cloud system.

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