Download PDFOpen PDF in browserUncertainty Estimations in Different Components of a Hybrid ANN - Fuzzy - Kriging Model for Water Table Level Simulation9 pages•Published: September 20, 2018AbstractThe purpose of this study is to examine the uncertainty of a combined artificial neural network (ANN), kriging and fuzzy logic methodology, which can be used for spatial and temporal simulation of hydraulic head in an aquifer. This methodology was applied in the past, while the verification of the model was performed by implementing it in a new study area, in Miami – Dade County, FL, USA. The percentile methodology was applied as a first approach in order to define the ANN uncertainty. As a second approach, the uncertainty of the ANN training is tested through a Monte Carlo procedure. The model was executed 300 times using different training set and initial random values each time. The training results constituted a sensitivity analysis of the ANN training to the kriging part of the algorithm. The training and testing error intervals for the ANNs and the kriging prediction intervals calculated through this procedure can be considered narrow compared to the complexity of the study area. For the third and final approach used in this work, the uncertainty of kriging parameter was calculated through the Bayesian kriging methodology. The results derived can prove that the simulation algorithm can provide consistent and accurate results.Keyphrases: Artificial Neural Networks, bayesian uncertainty, Fuzzy Logic, Kriging In: Goffredo La Loggia, Gabriele Freni, Valeria Puleo and Mauro De Marchis (editors). HIC 2018. 13th International Conference on Hydroinformatics, vol 3, pages 2042--2050
|
