The Research Of Self-organizing Data Mining Based On GMDH Principle

Data mining is a process of extracting novel and useful knowledge from large amounts of incomplete raw data. Most important for a more sophisticated data mining application is to limit the involvement of users in the general data mining process to the inclusion of existing a priori knowledge while making this process more automated and more objective.Self-organization is a process during which a system, when driven by its own inherent mechanism, develops from roughness to fine and improves its complexity and precision. The spacial, temporal or functional structure of a Self-organizing system forms only through the interaction of the system without any external interference.Self-organizing data mining introduces the self-organizing theory to the data mining process, applying the GMDH (Group Method Data Handling) principle to make the process more automated and more objective. An optimal complex model is created in a self-organizing modeling process and involvement of the users is reduced.This thesis expatiates on the theory of self-organizing data mining, including its main idea, basic pattern, and technology - GMDH Method.Applying the Self-organizing data mining technology, an analysis model is created automatically, which revealing the relation among the factors that influence national economy. A model created by GMDH is effective in the application of forecasting the current quantity of subway passengers.