MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202631123045 A) filed by Indian Institute Of Technology Indian School Of Mines, Dhanbad on October 14, 2026, for Hybrid Machine Learning And Geostatistical Method And System For Mineral Grade Prediction.
Inventors include Dr. Mrinal Kanti Mukherjee; Dr. Dipankar Ray; and Dr. B C Sarkar.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: A computer-implemented method and system (120) for predicting mineral grade in a subsurface geological domain are disclosed. Borehole data (402) comprising geological feature variables, observed mineral grade values, and corresponding three-dimensional spatial coordinates are received. A machine learning prediction (410) is generated using a machine learning model trained on the geological feature variables while excluding the three-dimensional spatial coordinates as input variables. Residual values (412) between observed and predicted mineral grade values are computed and subjected to variogram analysis (414) to characterize spatial dependence. Three-dimensional ordinary kriging (416) is performed on the residual values to generate kriged residuals (418) at unsampled spatial locations. A hybrid prediction (420) is generated by combining primary mineral grade predictions with corresponding kriged residuals (418), thereby generating spatially corrected mineral grade estimates.
Disclaimer: Curated by HT Syndication.
