MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641049547 A) filed by J. B. Institute Of Engineering And Technology, Hyderabad, Telangana, on April 18, for 'a system and method for adaptive machine learning in dynamic data environments with context-aware model evolution.'
Inventor(s) include Dr. Akheel Mohammed; Mr. Md Mahebub Ali; Mrs. Umme Hani Sara; Mrs. Radhika Reddy Devarapalli; Mrs. B. Sandhya Rani; and Ms. Intesar Fatima.
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: "The present invention discloses a system and method for adaptive machine learning in dynamic data environments, wherein machine learning models are continuously updated based on real-time data streams and contextual inputs. The system comprises a data ingestion module for collecting and preprocessing heterogeneous data, a context analysis engine for extracting relevant contextual parameters, and a drift detection module for identifying variations in data distribution. Upon detection of drift, a model evolution engine dynamically updates models using adaptive learning techniques without requiring complete retraining. A multi-model orchestration module evaluates and deploys optimal models based on performance metrics, while an explainability module provides interpretable insights into model decisions. The invention enhances predictive accuracy, reduces computational overhead, and ensures robust and scalable operation across diverse applications including finance, healthcare, cybersecurity, and intelligent automation systems."
Disclaimer: Curated by HT Syndication.
