A Fuzzy Decision Support System Integrating Neural Networks and Genetic Algorithms for Financial Planning and Management

Abstract

In this paper, a new financial management decision support system (FDSS) is proposed, which aims to improve the accuracy of budget management, the speed of decision and the ability of risk prediction by integrating fuzzy logic (FL), multilayer perceptron (MLP) and genetic algorithm (GA). The FDSS was designed to address the uncertainty and complexity of traditional approaches to financial data. Specifically, fuzzy logic is used to deal with fuzziness and uncertainty in financial data, MLP is used to capture complex nonlinear relations in data, and GA optimizes parameter settings in decision-making process, improving adaptability and robustness of system. Numerical experiments are carried out to verify the effectiveness of the FDSS. The results show that FDSS has achieved significant improvements in budget management, reducing budget overruns from 15% to 5%, reducing budget preparation time from 30 days to 15 days, and improving budget adjustment accuracy from 70% to 90%. In terms of risk early warning ability, the non-warning rate of market risk, credit risk and liquidity risk decreased by 75%,80% and 80% respectively, and the advance time of early warning increased by 7 to 10 days on average. In addition, the investment decision-making period was reduced from 45 days to 25 days, the deviation of return on investment was reduced from ±10% to ±5%, and the accuracy of screening invalid investment projects was improved from 70% to 90%. The overall financial health index improved from 7.5/20 to 16/20, indicating a significant improvement in the overall financial position of the institution. Experimental results show that FDSS is superior to the traditional methods in data processing speed, budget prediction accuracy, risk warning accuracy and user satisfaction. Especially in terms of data processing speed, FDSS is reduced from 120 seconds to 30 seconds, which improves the efficiency by nearly four times. To sum up, the FDSS proposed in this paper shows significant advantages in improving financial management efficiency, accuracy and user satisfaction by combining advanced technologies such as fuzzy logic, MLP and GA. The system can not only effectively deal with the uncertainty in the financial environment, but also improve the quality of decision-making, and provide a more robust decision support tool for financial institutions and enterprises.