MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641095134 A) filed by Lords Institute Of Engineering And Technology on August 06, 2026, for Machine Learning Enhanced Mathematical Framework For Ocean Pollution Prediction Using Multi-Source Environmental Data Fusion.
Inventors include Afser Unnisa; Moksud Alam Mallik; Mohammed Tajuddin; Mohd Khalid; Shaik Mohammed Ali; Parvez Ahmed Dar; Saleem Yousuf Lone; Amtul Muneer; Parvez Moiz; and Shaik Aslam.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: A computer-implemented framework for predicting ocean pollution integrates mathematical preprocessing, spatial modelling, feature optimization and machine learning to analyse heterogeneous environmental datasets obtained from multiple sensing and information sources. Environmental observations are subjected to noise reduction, missing-value estimation, normalization, spatial interpolation and temporal synchronization. A monitored marine region is represented as a weighted graph of observation locations, spatial connections and pollutant transport weights. Optimized environmental variables are processed using one or more trained machine learning models to estimate pollution probability, pollution severity and confidence for future time intervals. A pollution risk index combines contaminant concentration, transport factor, sensor-derived environmental score and weather influence using calibrated coefficients. The framework generates pollution hotspot maps, forecast timelines, alert notificati
Disclaimer: Curated by HT Syndication.
