MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641051240 A) filed by Sr University, Warangal, Telangana, on April 22, for 'system and method for ensuring data privacy and security using differential privacy and federated learning.'
Inventor(s) include Veeramalla Nikitha; and Dr P. Praveen.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "A system (100) and method for ensuring data privacy and security using differential privacy and federated learning is disclosed, wherein a plurality of client nodes (102) perform local model training (104) on private datasets without sharing raw data, and model updates generated at each node are processed using differential privacy module (108) to introduce controlled noise, thereby preventing identification of individual data records. the privacy-preserved (106) updates are transmitted to a central aggregation server (110) which combines the updates to generate a global model (112), and the global model (112) is redistributed to the client nodes (102) for iterative refinement until a desired performance level is achieved, enabling secure and decentralized data processing while maintaining data confidentiality and reducing the risk of data leakage across distributed environments such as healthcare, finance, and digital platforms."
Disclaimer: Curated by HT Syndication.
