MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202641007796 A) filed by Rahul Suryodai; and Desidi Narsimha Reddy, Mahabubnagar, Telangana, on Jan. 27, for 'machine learning for cybersecurity: detecting and preventing threats in the age of ai.'

Inventor(s) include Rahul Suryodai; and Desidi Narsimha Reddy.

The application for the patent was published on Feb. 13, under issue no. 07/2026.

According to the abstract released by the Intellectual Property India: "In the era of AI, cybersecurity faces unprecedented challenges from sophisticated, adaptive threats. This invention discloses a comprehensive machine learning framework for detecting and preventing such threats. The system architecture includes data ingestion modules that collect network traffic, system logs, and behavioral data. Preprocessing employs techniques like PCA for dimensionality reduction, ensuring efficient handling of large datasets. The core innovation is a hybrid ML ensemble: CNNs analyze spatial patterns in data for anomaly detection, RNNs process temporal sequences to predict attack progressions, and reinforcement learning agents dynamically select and refine preventive actions. Adversarial training is integrated to counter AI-based attack evasions, enhancing model resilience. Upon threat identification, the system autonomously executes responses, such as quarantining endpoints or updating access policies, minimizing human intervention and response time. Federated learning enables collaborative model improvement across decentralized networks while preserving data privacy. Evaluation through simulations and real-world case studies demonstrates superior performance: 98% detection accuracy, 5% false positive rate, and 70% reduction in breach impacts compared to traditional methods. A numerical example on a 1,000-endpoint network shows rapid ransomware mitigation, preventing widespread damage. This invention's novelty lies in its proactive, AI-resilient approach, addressing gaps in prior arts focused on reactive detection. Applicable to enterprises, critical infrastructure, and emerging technologies, it fortifies digital ecosystems against evolving cyber risks."

Disclaimer: Curated by HT Syndication.