Vorrichtung und Verfahren zum Trainieren eines Neuronalen Netzes

Details

Veröffentlichungs-Nr.

EP4749523

Anmeldetag

25. November 2024

Veröffentlichung

27. Mai 2026

Rechtsraum

EP

IPC

G06N3/045, G06N3/0985

Abstract

Computer-implemented method for training a neural network (60), wherein the neural network is configured to accept an input (x) characterizing at least one sensor measurement (S) and provide an output (y) characterizing a classification and/or regressions result of the at least one sensor measurement (S) and/or a probability of the sensor measurement (S) to occur among a set of sensor measurements or wherein the neural network (60) is configured to provide an output (y) characterizing a prediction of a sensor measurement for creating a training and/or test dataset for training another machine learning system, wherein the method for training comprises the steps of:• Performing a singular value decomposition of a weight matrix or a weight tensor of the neural network (60);• Grouping singular values and corresponding singular vectors into a plurality of groups based on the magnitudes of the singular values;• Training the neural network (60) by adapting the singular values and/or singular vectors of only the group corresponding to the lowest singular values or adapting the singular values and/or singular vectors of only a predefined number of groups corresponding to the lowest singular values.

Anmelder

Firma

Robert Bosch GmbH

Land

🇩🇪 Deutschland