MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202611070143 A) filed by Noida International University on June 04, 2026, for Federated Continual Learning System For Privacy-Preserving Model Training.
Inventors include Aditi Gaur; Archna Goyal; Manish Kumar; and Rahul Kumar Singh.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The invention relates to privacy-preserving machine learning and, particularly, to federated and continual learning for model training. The invention addresses the problem of training machine learning models in distributed environments while reducing centralisation of raw local data and permitting model improvement over time. A coordinator and distributed client nodes perform federated learning, wherein each client node conducts local learning using retained raw local data. Model update information is used to update a global model. The learning process is repeated through successive learning cycles to provide continual learning. The invention is useful for computing systems, distributed artificial intelligence platforms, cloud-edge networks, and organisations requiring privacy-preserving and adaptive model training.
Disclaimer: Curated by HT Syndication.
