The Application Of Data Mining On Earthquake Forecast

In recent years,many people in information industry attach more importance to the data mining technique,which is attributed to the necessary result from the contradictory movement between the increasingly rapid growth of data volume and the relatively poor information. Systematic and in-depth, comprehensive, detailed study of the data mining technique is objectively required by the development of the global information technology, and must also stimulate earthquake analysis prediction.At present, data processing methods in the prediction of earthquake are still statistical methods. Data mining techniques are seldom used in the earthquake. Earthquake data is mass, nonlinear, high dimensional, non-perfect, and noisy, so those analyzing and predicting methods very difficult get right conclusion. Data mining technique can process such data. Thus data mining technique has good prospect in earthquake broadcasting. The status of earthquake broadcasting is analyzed in detail and the relative problems, especially relation rules data mining technique, are discussed theoretically and technically based on data mining theory.Based on the basic concepts of the association rules,this dissertation analyses and studies the Apriori algorithms of association rules in detail and also depicts well the current various techniques of improving the efficiency of the algorithm. On the basis of the above, it further proposes the Aprior excavation algorithm of riddling compression, and effectively resolved the bottle drive of operating inefficiently and occupying the large memory space of the Apriori algorithms. It finally applied the algorithm to earthquake forecasts and conducted simulation experiments. The result from comparison demonstrated that the Aprior excavation algorithm of riddling compression greatly enhanced the efficiency to the mining of large database and the accuracy of the earthquake forecasts.