Patent · US Active

Deep reinforcement learning with fast updating recurrent neural networks and slow updating recurrent neural networks

US10872293B2 · kind B2 · utility

6Cited by

0References

30Claims

0Family size

Assignee

DeepMind Technologies Limited · GB

Inventors

Iain Robert Dunning · New York, US
Wojciech Czarnecki · London, GB
Maxwell Elliot Jaderberg · London, GB

Key dates

Filing date	May 29, 2019
Grant date	Dec 22, 2020
Priority date	—
Expiry date	May 29, 2039

Classification

Technology area (CPC G)Physics
CPC primaryG06N3/092
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for reinforcement learning. One of the methods includes selecting an action to be performed by the agent using both a slow updating recurrent neural network and a fast updating recurrent neural network that receives a fast updating input that includes the hidden state of the slow updating recurrent neural network.

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