Modeling, Reproduction, and Mapping of Geological Fields with and Without Measurement Noise. Part 2. Variational Modeling, Interpolation, and Smart Computation Methods

Resumo

Various methods for interpolation, smoothing, and neural network learning are examined. It is found that none of these methods is universal and absolutely reliable. Hybridization of these methods is recommended. Results are shown from a fuzzy neural network simulation with random location of the learning points and an unknown form of the sample distribution.