Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020142620 - MULTI-FORECAST NETWORKS

Publication Number WO/2020/142620
Publication Date 09.07.2020
International Application No. PCT/US2020/012073
International Filing Date 02.01.2020
IPC
G06N 5/04 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
5Computer systems using knowledge-based models
04Inference methods or devices
G06N 3/08 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
CPC
G06N 3/008
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
004Artificial life, i.e. computers simulating life
008based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. robots replicating pets or humans in their appearance or behavior
G06N 3/04
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architectures, e.g. interconnection topology
G06N 3/08
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
G06N 5/02
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
5Computer systems using knowledge-based models
02Knowledge representation
Applicants
  • SONY CORPORATION OF AMERICA [US]/[US]
  • SONY GROUP CORPORATION [JP]/[JP]
Inventors
  • CAPOBIANCO, Roberto
  • KOMPELLA, Varun
  • SUBRAMANIAN, Kaushik
  • MACGLASHAN, James
  • WURMAN, Peter
  • BAVEJA, Satinder
Agents
  • LIN, Vic
Priority Data
62/788,33904.01.2019US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) MULTI-FORECAST NETWORKS
(FR) RÉSEAUX À PRÉVISIONS MULTIPLES
Abstract
(EN)
A method and system for training and/or operating an artificial intelligent agent can use multi-input and/or multi-forecast networks. Multi-forecasts are computational constructs, typically, but not necessarily, neural networks, whose shared network weights can be used to compute multiple related forecasts. This allows for more efficient training, in terms of the amount of data and/or experience needed, and in some instances, for more efficient computation of those forecasts. There are several related and sometimes composable approaches to multi-forecast networks.
(FR)
L'invention concerne un procédé et un système d'entraînement et/ou d'exploitation d'un agent intelligent artificiel, qui peuvent utiliser des réseaux à entrées multiples et/ou à prévisions multiples. Des prévisions multiples sont des constructions de calcul, typiquement mais pas nécessairement des réseaux neuronaux, dont les poids de réseau partagés peuvent être utilisés pour calculer de multiples prévisions associées. Cela permet un entraînement plus efficace en ce qui concerne la quantité de données et/ou d'expérience nécessaires, et dans certains cas, un calcul plus efficace de ces prévisions. Il existe plusieurs approches associées et parfois composables de réseaux à prévisions multiples.
Also published as
Latest bibliographic data on file with the International Bureau