Data and Model Uncertainty

Terminology (uncertainty, ambiguity, vagueness, fuziness, quality, accuracy, errors, reliability), semantic issues. Dominant concepts in dealing with uncertainty (inheritent complexity and details of the world and phenomena, inheritent vagueness of definitions and concept, missing natural units for analysis, ambiguity of indirect indicators). Sources of errors and uncertainty. Introduction to application of Monte Carlo method. Geographical uncertainty (crisp boundaries, location etc.), attribute uncertainty. Ecological falacy, MAUP and data agreggation. Spatial autocorrelation. Errors of vector-raster conversions. Error propagation (statistical aproach, simulation aproach). Error balancing. Internal and external validation. Senstivity analysis. Methods based on simulations or decomposing the variance of the output. Reliability and Survival in Econometrics and Finance. Metadata. Bayesian theory, Bayesian belief networks. Dempster-Shafer theory. Techniques for reducing, quantifying and visually representing uncertainty. Cost and benefits of uncertainty decreasing. Uncertainty of decision making.