Product Cost Accounting And Prediction Based On BP Neural Network

These years,with the development of industry and information in our country,the huge change has happened from production way to management way.The construction of product cost has changed tremendously.Therefore,the traditional management way can not describe the cost and provide the exact accounting information for the further decision.The research is from the two aspects of the product cost and the cost prediction.The prediction of product cost is based on business accounting.The further forecasting is based on the product cost accounting,so we’d better do well in product cost accounting.If we can complete the forecasting well,we can make excellent further decisions.In this paper,the combination of the traditional cost accounting method and linear regression method is used to predict the cost of the product;the combination of activity based costing and multiple linear regression and BP neural network is used to predict the cost of the product.On the basis of the theory,using the data of Y,a chemical company,complete the cost forecast of glycerol.The accuracy of BP neural network is higher than 90%.Because the cost data is generally nonlinear,BP neural network has the universal applicability in cost prediction.