Based On The SVR Of Coal Mine Safety Cost Prediction Research

Coal industry is one of the high-risk industries in China.Coal is one of the most important resources of human production and life.Its output and consumption are the first of all kinds of resources.The situation of coal industry safety production is extremely serious,coal mine safety accidents will still occur,and the national and people's economy will suffer form lossing.The safety investment of coal mine is not enough,and the safety investment is not enough,which is the important reason for the high frequency of coal mine safety accidents.In this paper,those factors affecting the safety cost of coal mine enterprises are analyzed,and the index system of the safety cost of coal enterprises is given from the point of view of the safety cost function of the coal mine enterprises.The grey correlation degree and the theoretic al knowledge of support vector machines in the grey system are introduced,and the grey correlation analysis is used to analyze the safety cost of the coal mine.The main factors that affect the cost of coal mine safety are extracted from the index system,and the index system of prediction model is established.In order to realize the automatic optimization and selection of the parameters,the particle swarm optimization,genetic algorithm and cross validation are used in the parameter optimization of the support vector machine prediction model,and the results are compared.First of all,the cost of coal mine safety management,data mining technology,data mining in coal mine in the research,data mining in the aspects of cost management research literature at home and abroad are reviewed.The concept,task,process and method of data mining are introduced respectively.The input data of coal mine safety cost are used as training samples and inspection samples.The support vector regression machine is used in this paper to predict and compare the coal mine safety cost respectively.The three optimization algorithms can effectively select the parameters of the coal mine safety cost prediction model of the support vector machine.Compared with the cross validation and genetic algorithm,the optimization operation process of the particle swarm optimization is relatively simple.Single,model training time is short,classification precision is high,and it has good generalization performance.The corresponding optimization scheme is proposed from the safety cost configuration and the safety cost condition.