MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123028 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 5, 2025, for 'machine learning system for real-time predictive analytics.'
Inventor(s) include Dr. Srinivas Koppu; and Kim Sharma.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a machine learning system for real-time predictive analytics. The system includes a data ingestion layer configured to receive data from multiple sources including financial data streams, cybersecurity logs, and industrial sensor data, a data processing pipeline configured to process the received data through extract-transform-load operations, a model management layer configured to train and store machine learning models including Long Short-Term Memory networks, autoencoders, and ensemble methods, an inference layer configured to deploy the machine learning models for real-time prediction generation across multiple domains including financial forecasting, cybersecurity anomaly detection, and predictive maintenance. The method (100) includes ingesting (101) data from multiple sources, processing (102) the ingested data including data validation and preprocessing operations, training (106) machine learning models using the processed data, and deploying (108) the trained models to inference servers for real-time prediction generation."
Disclaimer: Curated by HT Syndication.
