A fuzzy logic based explicit declustering technique

During the spatial estimation of geoscience resource variables, quantity or quality minerals and hydrocarbons can be represented by a broad range properties, including geochemical, geotechnical, other physical measures. Preferential sampling within region interest causes biased global parameters due to clustered patterns. Unbiased sample distribution is essential for conducting conditional simulations model uncertainty spatially distributed attributes, e.g. geochemical content metal porosity. Therefore, declustering procedures are applied during estimate an unbiased statistical measured variables. Traditional techniques such as cell do not consider grade clustering, i.e., similarity variables neighbourhood. This paper presents technique that explicitly accounts clustering samples' attributes these samples. In proposed method, samples were first classified into (grade) clusters using Fuzzy c-means algorithm. Declustering weights derived Mamdani based Inference System various T-norm operations. The was publicly available GSLib Walker Lake datasets. It shown scheme produced more accurate results than those obtained with traditional technique.