Optimization and innovation of enterprise finance and accounting supervision system under big data technology

Abstract

Abstract With the continuous development of the social economy, financial and accounting risk control and early warning have become an important part of the sustainable development of enterprises. This paper combines the C4.5 decision tree and Benford law-based random forest audit warning model by constructing enterprise financial risk assessment indicators and audit warning indicators, calculates the indicator data of 100 companies to get the financial risk assessment rule set, and validates it with the financial data of Company A in 2018-2020 as a sample. Our method of obtaining the audit warning interval for 8 indicators and validating it is by using Company B’s indicator data from 2019-2020 as a sample. The assessment results are ‘yes’ when company A is used as an example for empirical analysis, confirming the accuracy of the financial risk assessment model. Early warning intervals are obtained from the Random Forest audit early warning model, in which accounts receivable ledger balance X1 > 5.72, accounts receivable aging X7 > 33.14, accounts payable aging X8 > 4.76, and provision for bad debts X9 > 14.10. The result of the test in the fourth quarter of 2019 for Company B is an early warning status with a probability of 73%. The warning interval is triggered by four indicators, which include the accounts receivable ledger balance X1, accounts receivable aging X7, accounts payable aging X8, and bad debt provision X9.