Intellectual Property India Publishes Patent Application for 'Dynamic Ensemble Learning System For Adaptive Model Fusion In AI Applications' Filed by Vardhaman College Of Engineering

MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123345 A) filed by Vardhaman College Of Engineering, Hyderabad, Telangana, on Dec. 7, 2025, for 'dynamic ensemble learning system for adaptive model fusion in ai applications.'

Inventor(s) include Mrs. Yadla Sunanda; Mr. P Vijaya Raghavulu; Ms. Swati Singh; Mr. K Venkatesh; Mr. Shobanbabu R J; and Mr. Mohd Salahuddin.

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 invention provides a Dynamic Ensemble Learning System for adaptive model fusion in AI applications, designed to improve accuracy, reliability, and responsiveness under changing data conditions. The system integrates multiple heterogeneous machine learning models and continuously evaluates their performance using real-time metrics such as accuracy, drift, confidence, and latency. An adaptive fusion engine dynamically selects, reweights, or excludes models based on current performance, while a feedback-control loop ensures continuous optimization by recalibrating fusion strategies in response to data distribution shifts. The system enables robust, scalable, and intelligent decision-making across diverse AI domains by using adaptive strategies including weighted fusion, gating, and meta learning-based optimization."

Disclaimer: Curated by HT Syndication.