Research on Venture Capital Sum Model and loan Credit Evaluation Model based on Ant Colony algorithm
Download as PDF
DOI: 10.23977/FMESS2021012
Author(s)
Rui Chen, Ying Chen, Tianyu Zhang
Corresponding Author
Rui Chen
ABSTRACT
There is a lack of research on the application of personal credit risk assessment model in online lending and credit information industry, so it is particularly urgent to study the domestic P2P online loan industry which has many chaotic phenomena. Aiming at the problem of risk assessment of investment projects, a risk assessment method based on ant colony-analytic hierarchy process is proposed. under the condition of meeting the consistency requirement, the minimum difference of judgment matrix before and after adjustment is taken as the objective function, and aiming at the problems of weak global searching ability and low convergence accuracy of genetic algorithm (GA), the normal distribution crossover operator is introduced into the crossover operation of genetic algorithm, and is used to measure risk by CVaR. The adjustment problem of judgment matrix is transformed into the traveling salesman problem of ant colony algorithm, and the risk weight vector is calculated. The effectiveness of the method is verified by numerical examples, and the corresponding preventive measures are put forward for the main risks. Firstly, the personal credit evaluation model is established by using support vector machine, and the genetic algorithm is introduced to optimize the parameters of the model, and then the validity analysis and generalization analysis are made for the samples of two P2P network loan platforms. and according to the empirical results to explore the potential risks of credit brushing behavior. The empirical results show that GA-SVM model can effectively solve the problem of personal credit evaluation of P2P network loan platform, and has good robustness and promotion. The results show that when the method is applied to the risk assessment of overseas investment projects, a more reasonable weight vector can be obtained, which is helpful to formulate corresponding risk measures according to the potential main risks.
KEYWORDS
Improved genetic algorithm, portfolio, numerical simulation, Analytic hierarchy process