MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202611060031 A) filed by Ihab Mahdi Ali Almaameri; and Dr. Rohit Sharma on May 12, 2026, for System And Method For Continual Learning Using Memory-Augmented Neural Networks With Drift Adaptation.
Inventors include Ihab Mahdi Ali Almaameri; and Dr. Rohit Sharma.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: A system and method for continual learning using memory-augmented neural networks with drift adaptation is disclosed. The system comprises a data acquisition unit configured to receive sequential multidimensional data streams from sensing devices, communication interfaces, and distributed computing nodes. A preprocessing processor performs signal normalization, temporal synchronization, feature scaling, and encoded feature vector generation for the received data streams. A neural computation processor comprising layered neural processing arrays generates adaptive contextual feature embeddings associated with evolving operational conditions. A memory augmentation unit including episodic memory storage portions, contextual representation storage portions, and long-term knowledge retention storage portions selectively stores and retrieves feature embeddings based on contextual significance, temporal recurrence frequency, and predictive relevance. A drift detection processor continuously compares current feature distribution characteristics with historical distribution characteristics to identify concept drift conditions.
Disclaimer: Curated by HT Syndication.
