MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641071893 A) filed by Sr University on June 10, 2026, for Intelligent Iomt Cyber Attacks Detection Through Prior-Data Fitted Networks Driven Explainable Generalized Additive Modelling.
Inventors include G Indira Priyadarshini; and Dr. N. Venkatesh.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: The present invention relates to an intelligent and interpretable intrusion detection system for securing Internet of Medical Things environments against evolving cyber threats. The system utilizes a Tabular Prior-Data Fitted Network (TPDFN) for extracting high-dimensional features from IoMT network traffic by leveraging meta- learning across diverse tabular data distributions to capture latent relationships while reducing overfitting and computational complexity. The extracted features are processed using a Random Tree Generalized Additive Model (RTGAM) for accurate multi-class classification of cyber attacks, combining nonlinear decision-making with additive interpretability. An explainability layer incorporating Local Interpretable Model-Agnostic Explanations (LIME) provides transparent, feature-level reasoning for each prediction. The system effectively classifies multiple IoMT traffic categories including benign and various attack types such as denial-of-service, distributed denial-of-service, reconnaissance, protocol-based intrusions, and spoofing, thereby enabling reliable, scalable, and explainable cybersecurity for modern healthcare infrastructures.
Disclaimer: Curated by HT Syndication.
