Research And Application Of Management Decision Analysis And Application Of Tiancheng Environmental Protection Enterprise Based On Big Data Technology

With the increasingly fierce competition among enterprises,how to quickly and accurately predict the market price of raw materials is of great significance to control the cost of enterprises and assist enterprises in purchasing decisions.The starting point of this paper is to solve the intelligent decision-making problem of raw material procurement of Tiancheng environmental protection enterprise.Taking the price of chemical fiber product polyester staple fiber as the research object,this paper selects the price data of polyester staple fiber from 2007 to 2020 as the data support,and uses data crawler,maximum expectation algorithm,correlation analysis and other technologies to standardize the data processing,Based on ARIMA model + LSTM model + BP neural network model,polyester staple fiber price forecasting is carried out.The optimal model is selected through model comparison and applied to polyester staple fiber price forecasting system,which realizes the function of selecting different models for price forecasting according to different data characteristics and assists Tiancheng environmental protection enterprise in purchasing decision.The research contents of this paper are as follows:(1)Data collection and processing.Through the web crawler technology to obtain polyester staple fiber price,output and other data,and the existing Tiancheng environmental protection enterprise database polyester staple fiber market price data for unified format processing,the missing part of the data is supplemented by EM algorithm,to obtain standardized and unified high-quality sample data.(2)Correlation analysis of influencing factors.In depth study and analysis of various factors that may affect the price of polyester staple fiber,Pearson coefficient is used for correlation analysis,and polyester staple fiber output,raw material price,export quantity and import price with higher correlation coefficient are selected for modeling analysis.(3)The selection and establishment of the model.The time series model,BP neural network model and LSTM model are selected to simulate and train the short fiber price,and the future price trend is predicted,and the prediction results are compared.Through comparative analysis,ARIMA model is suitable for polyester staple fiber price prediction with small fluctuation of original data and short prediction period,LSTM model is suitable for polyester staple fiber price prediction with large fluctuation of original data,and BP neural network is suitable for polyester staple fiber price prediction with nonlinear relationship of original data.Finally,the three models are applied to the polyester staple fiber price forecasting system,and different models are selected according to the different needs of the company.(4)Develop polyester staple fiber price prediction and analysis system,complete data management,price prediction and other modules,provide scientific basis for raw material procurement of Tiancheng environmental protection enterprise,and then reduce enterprise investment cost.