MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541123637 A) filed by Sri Shanmugha College Of Engineering And Technology, Salem, Tamil Nadu, on Dec. 8, 2025, for 'real-time data imputation system using self-supervised machine learning algorithms.'
Inventor(s) include Dr. S. Nithyakalyani; Dr V Suresh Babu; Dr. K. Meenambika; Dr V. K. Senthil Ragavan; Dr. G. Petchinathan; Ms. S. Manimala; Mr. V. Mouliraj; and Dr. B. Gopinath.
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: "This paper presents a Real-Time Data Imputation System Using Self-Supervised Machine Learning Algorithms, a novel framework that automatically reconstructs missing values within continuously streaming data through masked reconstruction pretext tasks and deep encoder-decoder architectures trained without manual annotations. Unlike traditional supervised imputation methods requiring labeled training pairs or statistical approaches relying on restrictive distributional assumptions, our system exploits the intrinsic structure of incomplete data via self-supervised learning objectives such as temporal denoising and masked value prediction, enabling robust representation learning across heterogeneous, high-dimensional, and non-stationary data streams. Evaluated across multivariate time series, tabular records, and sensor data, our approach demonstrates substantial improvements over existing batch-trained and statistical baselines while maintaining sub-second latency and scalability for integration into distributed data infrastructure, thereby enabling organizations to operate reliably on imperfect streaming data for critical forecasting, anomaly detection, and decision-support tasks."
Disclaimer: Curated by HT Syndication.
