Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020091865 - FINITE RANK DEEP KERNEL LEARNING FOR ROBUST TIME SERIES FORECASTING AND REGRESSION

Publication Number WO/2020/091865
Publication Date 07.05.2020
International Application No. PCT/US2019/043934
International Filing Date 29.07.2019
IPC
G06N 20/10 2019.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
10using kernel methods, e.g. support vector machines
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
G06K 9/6256
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
6217Design or setup of recognition systems and techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
6256Obtaining sets of training patterns; Bootstrap methods, e.g. bagging, boosting
G06N 20/10
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
10using kernel methods, e.g. support vector machines [SVM]
G06N 3/0481
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
0481Non-linear activation functions, e.g. sigmoids, thresholds
G06N 3/084
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
084Back-propagation
Applicants
  • INTUIT INC. [US]/[US]
Inventors
  • DASGUPTA, Sambarta
  • KUMAR, Sricharan
  • SRIVASTAVA, Ashok
Agents
  • PATTERSON, B. Todd
  • TRANSIER, Nicholas R.
Priority Data
16/212,60106.12.2018US
62/755,22902.11.2018US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) FINITE RANK DEEP KERNEL LEARNING FOR ROBUST TIME SERIES FORECASTING AND REGRESSION
(FR) APPRENTISSAGE DE NOYAU PROFOND DE RANG FINI POUR UNE PRÉVISION ET UNE RÉGRESSION DE SÉRIES CHRONOLOGIQUES ROBUSTES
Abstract
(EN)
Certain aspects of the present disclosure provide techniques for performing finite rank deep kernel learning. In one example, a method for performing finite rank deep kernel learning includes receiving a training dataset; forming a set of embeddings by subjecting the training data set to a deep neural network, forming, from the set of embeddings, a plurality of dot kernels; combining the plurality of dot kernels to form a composite kernel for a Gaussian process; receiving live data from an application; and predicting a plurality of values and a plurality of uncertainties associated with the plurality of values simultaneously using the composite kernel.
(FR)
Certains aspects de la présente invention concernent des techniques de réalisation d'une apprentissage de noyau profond de rang fini. Dans un exemple, un procédé pour réaliser un apprentissage de noyau profond de rang fini comprend la réception d'un ensemble de données d'apprentissage ; la formation d'un ensemble d'intégrations par soumission de l'ensemble de données d'apprentissage à un réseau neuronal profond, la formation, à partir de l'ensemble d'intégrations, d'une pluralité de noyaux de points ; la combinaison de la pluralité de noyaux de points pour former un noyau composite pour un processus gaussien ; la réception de données en direct à partir d'une application ; et la prédiction d'une pluralité de valeurs et d'une pluralité d'incertitudes associées à la pluralité de valeurs simultanément à l'aide du noyau composite.
Latest bibliographic data on file with the International Bureau