The Underground Fluid Data Correlation Analysis System Based On Association Rules

Earthquake prediction is the prediction about time,place and magnitude of the future earthquakes,is of great importance to national public security protection.With the rapid development of seismic observation technology,the scale of the seismic data has increased rapidly.How to find out the hidden rules before the earthquake is a key issue need to be solved urgently.The association rule is one of the main research contents in the field of data mining.It can effectively depict and describe the association rule between the underground fluid data and magnitudes before earthquakes,and is of great value to improve the accuracy of earthquake prediction.In this thesis,the association rule mining method is used to study the association rule between the underground fluid data and magnitudes.The main findings are as follows:(1)According to the characteristics of underground fluid data,such as large,high dimensional,continuous,nonlinear and so on,a weighted association rule mining algorithm is presented.This algorithm is used to discretize the underground fluid data,which can not only preserve the important information,such as the maximum and minimum values of the original underground fluid data,but also reduce the attribute dimension of the underground fluid data effectively.According to the weighted support and k-support expectations,the frequent patterns that contain characteristics of the underground fluid data and magnitudes are extracted.According to the confidence threshold,a weighted association rule is established to describe the relationship between characteristics of the underground fluid data and magnitudes.(2)Using Eclipse and Oracle as development tools,a prototype system of underground fluid data correlation analysis is implemented based on association rules.The running results of the prototype system show that the correlation between characteristics of the underground fluid data and the regional magnitudes can be found by the weighted association rule mining method,which can provide reference for the scientific decision of the earthquakeprediction.