Global Patent Index - EP 3899808 A1

EP 3899808 A1 20211027 - METHOD FOR TRAINING A NEURAL NETWORK

Title (en)

METHOD FOR TRAINING A NEURAL NETWORK

Title (de)

VERFAHREN ZUM TRAINIEREN EINES NEURONALEN NETZES

Title (fr)

PROCÉDÉ POUR ENTRAÎNER UN RÉSEAU NEURONAL

Publication

EP 3899808 A1 20211027 (DE)

Application

EP 19812975 A 20191128

Priority

  • DE 102018222347 A 20181219
  • EP 2019082837 W 20191128

Abstract (en)

[origin: WO2020126378A1] A computer-implemented method for training a neural network (60), which is configured, in particular, to classify physical measurement variables, wherein the neural network (60) is trained using a training data set (X), and for training purposes, pairs comprising an input signal (x) and an associated desired output signal (y T ) are drawn from the training data set (X), and wherein parameters (θ) of the neural network (60) are adapted according to an output signal (y) of said neural network (60) when the input signal (x) is supplied and according to the desired output signal (y T ), characterised in that the pairs are always drawn from the whole training data set (X).

IPC 8 full level

G06N 3/08 (2006.01)

CPC (source: EP KR US)

G06F 18/213 (2023.01 - EP KR US); G06F 18/241 (2023.01 - US); G06N 3/047 (2023.01 - KR); G06N 3/063 (2013.01 - KR); G06N 3/08 (2013.01 - US); G06N 3/084 (2013.01 - EP KR)

Designated contracting state (EPC)

AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

Designated extension state (EPC)

BA ME

DOCDB simple family (publication)

WO 2020126378 A1 20200625; CN 113243021 A 20210810; DE 102018222347 A1 20200625; EP 3899808 A1 20211027; JP 2022514886 A 20220216; JP 7137018 B2 20220913; KR 20210099149 A 20210811; TW 202105261 A 20210201; TW I845580 B 20240621; US 2021406684 A1 20211230

DOCDB simple family (application)

EP 2019082837 W 20191128; CN 201980084359 A 20191128; DE 102018222347 A 20181219; EP 19812975 A 20191128; JP 2021535840 A 20191128; KR 20217022763 A 20191128; TW 108146410 A 20191218; US 201917295434 A 20191128