Research on Big Data Risk Control Model of Venture Capital

Abstract

The two main characteristics of venture capital are high risk and participation in management. Risk identification and risk evaluation before investing, risk supervise and control are important process that affect the success of venture investment. First of all, a risk evaluation index system is constructed. Partial correlation analysis is used to explore the indicators that can significantly affect the success of a company's investment, and to provide suggestions for the types of risks that start-ups should focus on controlling during the startup period. Then the principal component analysis method and the Logistic regression analysis method are combined to predict the success rate of investment, which can make up for the deficiency of the Logistic model and improve the prediction accuracy rate. Then use the test set data to calculate the accuracy of the model, and conduct an Omnibus test of the model coefficients to verify the significance of the equation. Then the SE-DEA model is constructed to calculate and compare the efficiency values of enterprises that accept different levels of post-investment services, and test the robustness of the SE-DEA model by adjusting the input and output indicators. Then through the Mann-Whitney U test and analysis, it is concluded that if the post-investment service is to have a good effect, how VC should choose the degree of intervention according to the state of the enterprise. That is to say, which indicators of the enterprise can be optimized by VC intervention in management. Finally, the models are evaluated and future research prospects in related fields are proposed.