ABSTRACT
An approximate model, based on Bayesian law, for calculating identification probability of classification problem is presented. An encoding procedure for existing variables, in which the geologist’s adventurous psychology and the credibility of the original data can be considered, is discussed in order to integrate spatial data. A pseudo-entropy criterion is proposed in order to select optimally drilling locations in oil-gas exploration. A case study based on the modified data of a gasfield is included.
