Intellectual Property India Publishes Patent Application for Real-Time Predictive Analytics System Using Federated Machine Learning For Privacy-Preserving Data Processing

MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641105556 A) filed by Mr. Win Mathew John; Naveen Parameshwarappa; Dr. Girish Dutt Gautam; Dr. Ayesha Riyajuddin Mujawar; Dr. Chandan Das; Samrat Bandyopadhyay; Dr. J. Sathiamoorthy; Dr. Sarika Keswani; Dr. Suvarna Mahavir Patil; Prof. Twinkle Gaur; and Prof. Dr. Harikumar Pallathadka on September 02, 2026, for Real-Time Predictive Analytics System Using Federated Machine Learning For Privacy-Preserving Data Processing.

Inventors include Mr. Win Mathew John; Naveen Parameshwarappa; Dr. Girish Dutt Gautam; Dr. Ayesha Riyajuddin Mujawar; Dr. Chandan Das; Samrat Bandyopadhyay; Dr. J. Sathiamoorthy; Dr. Sarika Keswani; Dr. Suvarna Mahavir Patil; Prof. Twinkle Gaur; and Prof. Dr. Harikumar Pallathadka.

The application for the patent was published on September 11, 2026, under issue no. 37/2026.

Abstract: The present invention provides a real-time predictive analytics system employing federated machine learning for privacy-preserving processing of distributed datasets. The system comprises a plurality of distributed data nodes configured to locally acquire and preprocess data, train respective predictive models, and generate model updates without transferring raw data to a centralized environment. A privacy management module applies privacy-preserving operations to the model updates, while a secure communication module transmits the protected updates to a federated coordination environment. An update validation module evaluates received updates, and an adaptive aggregation module combines validated updates to generate an updated global predictive model. A real-time inference module generates predictive outputs using the global model, while monitoring and feedback modules evaluate model performance, data distribution changes, prediction errors, and model drift to initiate subsequent federated learning operations. Thus, the invention enables continuous predictive analytics while maintaining privacy of distributed source datasets.

Disclaimer: Curated by HT Syndication.