Patent · US Active

Method and device for training a machine learning routine for controlling a technical system

US12361280B2 · kind B2 · utility

0Cited by

0References

20Claims

0Family size

Assignee

Siemens Healthcare Diagnostics Inc. · US

Inventors

Markus Michael Geipel · Munich, DE
Stefan Depeweg · Munich, DE
Christoph Tietz · Ottobrunn, DE
Gaby Marquardt · Eschborn, DE
Daniela Seidel · Baiersdorf, DE

Key dates

Filing date	Jul 29, 2019
Grant date	Jul 15, 2025
Priority date	—
Expiry date	Nov 11, 2042

Classification

Technology area (CPC G)Physics
CPC primaryG06N20/20
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

To train a machine learning routine (BNN), a sequence of first training data (PIC) is read in through the machine learning routine. The machine learning routine is trained using the first training data, wherein a plurality of learning parameters (LP) of the machine learning routine is set by the training. Furthermore, a value distribution (VLP) of the learning parameters, which occurs during the training, is determined and a continuation signal (CN) is generated on the basis of the determined value distribution of the learning parameters. Depending on the continuation signal, the training is then continued with a further sequence of the first training data or other training data (PIC2) are requested for the training.

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