Multiple Regressive Pattern Recognition Technique: An Adapted Approach for Improved Georesource Estimation
Multiple Regressive Pattern Recognition Technique (MRPRT) is an adapted approach for improved geologic resource estimation. We developed and tested this approach for the Platinum (Pt) bearing region near Goodnews Bay, Alaska, which presents an example of a complex depositional environment. We applied geospatial and pattern recognition methods to assess the spatial distribution of offshore Pt in the Goodnews Bay area from point data collected by various agencies. We used the coefficient of correlation (r) and the Nash-Sutcliffe efficiency (E) to quantitatively assess the degree of accuracy of the estimated Pt distribution. We split the study area, based on trend analysis, into two regions: inside the Bay and outside the Bay. We could not obtain appreciable estimates from the geospatial and pattern recognition methods. Using MRPRT, we were able to improve r from 0.57 to 0.93 and the E from 28.31 to 92.91 inside the Bay. We achieved improvement in r from 0.55 to 0.61 and E from 28.46 to 34.52 outside the Bay. The reasons for a non-significant improvement outside the Bay have been discussed. The results indicate that the proposed MRPRT has wide application potential in georesource estimation where input data is often scarce. © 2011 International Association for Mathematical Geology.
Natural Resources Research
Multiple Regressive Pattern Recognition Technique: An Adapted Approach for Improved Georesource Estimation.
Natural Resources Research,
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/4972