Data Mining With Weka Tools In The Application Of The River Channel Flood Forecasting

With the development of Hydrological Sciences, hydrological automaticobservation instruments were applied, people recorded a large number of hydrologicalobservation data. In the hydrology research and application, people usually put on dataclassification storage to the hydrological database, including the rainfall database, waterdatabase and sediment database. The hydrological database has a good support to dealwith the daily affairs. With the hydrological data in the database is increasing,hydrological workers hoping to find a deeper connection from the mass of hydrologicaldata, the data mining technology can solve this problem. Hydrological workers facehydrological data mining research to become a new topic.In this article, first, the existing data mining technology and hydrologicalforecasting model is briefly introduced, analysis of the hydrological data miningapplication characteristic and development trend. Then introduces the concepts of datawarehouse, the existing database and data warehouse are compared. From the systemstructure of data warehouse and data model has made some reviews. According to thedifferent themes for the establishment of data warehouse, describe level characteristic ofthe hydrological data mining system model. The applications of data warehouse in theYellow River basin are discussed. Next, the introduction of data mining tools Weka’sbackground and function, application of Weka in regression analysis algorithm,combined with the actual observation data to establish prediction model, contrast thepredicted result and the actual data in the picture. Case of the prediction model foranalysis, get the effect of the prediction model under various circumstances, determinethe forecasting model of applicable conditions, makes the forecast model has practicalvalue. Finally, for the next study on how to improve the prediction accuracy, gives someopinions and suggestions.