Non-linear univariate and multivariate spatial modelling and optimal design

摘要

This thesis developed a novel adaptive methodology for the optimal design of additional sampling based on a geostatistical model that can preserve both multivariate non-linearity and spatial non-linearity present in spatial variables. This methodology can be applied in mining or any other field that deals with spatial data. The results from the different environment case studies demonstrated the potential of the proposed design methodology.