The Application Of Data Mining Technology In Meteorological Forecasting

Aiming at the problem of the global frequent extreme weather conditions,it is very important to discover the appearance of the disaster weather.With the era of big data's coming,data mining technology is applied to the weather forecast to analysis the correlation between various meteorological factors and improves the accuracy of weather forecasts,it has a very important practical significance.The meteorological forecast is based on traditional statistical prediction model,the probability of relevant methods in the field of the historical data to establish one or more model.However,the traditional statistical methods often applied to a large number of forecast objects,the more forecast objects,the higher precision rate on finding out the relationship between the forecast factors and forecast object.However,in application,it is normal to see that a particular type of weather forecast is a very common.Using data mining technology can target a specific weather object,quickly handle huge amounts of weather data,excavate potential.it is hard to find the correlation of the prediction factors,which will has a greatly useful improve at the accuracy of weather forecasts.Nowadays,there are more and more scholars and researchers pay close attention to the using of intelligent algorithms in the field of data mining application.Gravitation Move Algorithm(GMA)which is presented this year,as a kind of heuristic optimization algorithm,compared with the traditional Particle Swarm Optimization(PSO)algorithm which has greatly improved performance,but still has its drawbacks.To improve the searching performance of gravitation move algorithm,in accordance with problems of bad performance in search accuracy and slow convergence speed in the high dimensional space optimization,PGMA is proposed by introducing affinity to improve the algorithm convergence and search precision,and this improved Gravitational Move Algorithm(GMA)changes the particle ' s gravitational force calculation formula.It includes the principles of gravitational move algorithm and the structure of the affinity,and the affinity for the appropriate transformation is added to the formula resultant force.Then the formula resultant force is modified.Thirteen benchmarks function are tested and show that new algorithm is better than GMA with both a steady convergence and a better accuracy of solution.Then the PGMA algorithm is applied to the field of association rule mining,and through the experiment to prove its performance in the field of association rule mining is improved.The Test for Independence-PGMA scheme of this kind of association rule mining and to the field of weather prediction,has the independence has been proved by the Shanghai meteorological data set improvement of TI-PGMA accuracy and effectiveness of association rules mining scheme.