A Comparative Study On Working Capital Risk Warning Of Listing Corporation In Manufacturing Industry

With an improvement of the national market economy, the industry of our country develops rapidly, the economic indicators increase with an exponential, especially the manufacturing industry has become one of the main support of GDP in China, China has become a manufacturing power, in the world also has the title of "world factory". However, due to the energy, environment, population and other constraints, Chinese economic growths lowed, investigate its reason, in response to the construction pace of innovation oriented country, China’s manufacturing industry gradually from the "world factory" to the advanced technology innovation industry excessive, enterprise transformation in the process of production and operation activities will face difficulties, especially the control and management of working capital and daily operation, it is necessary to put more attention to working capital operation due to improper financial risks caused. Therefore, in order to manufacturing industry steadily reform development, early warning research to manufacturing listing Corporation accurate working capital risk.Based on previous research on working capital risk, first of all, the theory basis for the research of working capital risk warning and data mining technology, the advantages and feasibility analysis of data mining method in the early warning analysis; secondly, to select two, A shares were warning processing manufacturing industry as the research object, and based on the selection criteria on the construction of index system, operating funds risk early warning index system, and a detailed description of data mining in the BP neural network, Logistic regression, decision tree C5.0three forecasting methods and the basic principles of model construction; finally, combined with the SPSS Clementine operating procedures, operating funds risk early warning empirical analysis to the selected36manufacturing industry Sample Firms, and model accuracy evaluation of the three kinds of forecasting methods, the article draws a conclusion. Through the research and analysis that, the listing Corporation operating funds risk early warning model based on data mining has the very strong warning capability; and three kinds of forecast model more close to the warning processing period, the prediction accuracy is high, show that the listing Corporation working capital risk is a dynamic variable, early warning model also has strong timeliness; by comparison analysis of longitudinal concludes, model of BP neural network prediction model established in this research is best, Logistic regression model is the worst, the middle C5.0decision tree model; data mining method, knowledge discovery, as the theoretical foundation of the model has high prediction precision, better than the statistical model based. Therefore the data mining technology is feasible in the analysis of operating funds risk early warning, the enterprises can fully application of data mining technology, to working capital risk management decision, to improve efficiency in the use of funds.