ARTIFICIAL-INTELLIGENCE METHODS IN GEOLOGICAL FORECASTING

The objective of applications of expert systems to geological forecasting is to use heuristic knowledge of experienced geologists, as well as their intuition and facility for drawing the right forecast conclusions based on incomplete data. At the same time, considerable experience is accumulated in geological forecasting with formal mathematical models that generalize the setting of some forecast problem classes. Thus, the "bottom-up" heuristic approach and the more formal "top-down" methods are aspiring to the same aim, so they supplement and strengthen each other. A forecast expert system called GENESIS that allows both approaches to be used has been produced. The heuristic part of GENESIS is the "empty" expert system, which consists of the knowledge definition language, destined for an expert geologist, the knowledge base used for forecast models storing, then forming knowledge base facilities, relational database for storing user's data and forecast specimen, components that organize the certain inference and its explanation.