US Patent Issued to Amazon Technologies on Sept. 23 for "Concept shift detection and correction using probabilistic models and learned feature representations" (German Inventors)

ALEXANDRIA, Va., Sept. 23 -- United States Patent no. 12,423,960, issued on Sept. 23, was assigned to Amazon Technologies Inc. (Seattle).

"Concept shift detection and correction using probabilistic models and learned feature representations" was invented by Lukas Stefan Balles (Berlin), Giovanni Zappella (Berlin) and Cedric Philippe Archambeau (Berlin).

According to the abstract* released by the U.S. Patent & Trademark Office: "Techniques for concept shift detection and correction using probabilistic models and learned feature representations are described. A gaussian process model is trained using representations generated by a primary machine learning (ML) model for existing training data elements in a training memory. For a new batch o...