Meteorological Data Analysis Based On Attribute Reduction And Bayesian In Cloud Environment

Nowadays,it is the high-speed stage of the development of big data.The potential value knowledge in data has been sought after by many scientific researchers.It is of great significance to mine the potential rules in meteorological data by using data mining technology to improve the ability of meteorological forecast.In the current research,the traditional machine learning algorithm is only suitable for the analysis of small data volume.The application of traditional machine learning methods in large data analysis is often not satisfactory.The efficiency and accuracy of the algorithm can not achieve the expected results,and the cost of traditional methods is high.There are the characteristics of high dimensional and strong dependence among meteorological data,but the traditional Bayesian method assumes that attributes are antagonistic to each other,and the accuracy of this method is low.Cloud Computing has become a popular way of computing large data because of its low cost and fast computing speed,but it requires proper architecture,efficient distributed and intensive tasks.Therefore,the selection of appropriate machine learning algorithms can efficiently process and analyze large-scale data.Parallelism and computational efficiency are the challenges that big data computing needs to overcome.This paper analyses the problems of machine learning and meteorological data analysis,and designs the correlation analysis strategy of meteorological data in cloud environment from two aspects of parallelism and computing efficiency.Based on the meteorological data of meteorological data center of China Meteorological Bureau and environmental monitoring station of Nanning Environmental Protection Bureau,this paper designs a knowledge reduction algorithm and improved Bayesian method to learn meteorological data in cloud environment.Firstly,the idea of knowledge reduction is applied to reduce the dimensionality of data.On this basis,Bayesian method of attribute association based on MapReduce is designed to explore the correlation of meteorological factors.In the experiment,the atmospheric visibility was used as the decision-making condition,and the time performance and classification accuracy were used as the main indexes to evaluate the performance.The experimental results show that the proposed method has better operation time and classification accuracy than the traditional Bayesian method.In this paper,the correlation analysis scheme of meteorological data based on cloud environment has effectively solved the shortcomings of traditional meteorological data mining toa certain extent.In view of the low cost and high scalability of cloud computing,this solution has good application value.