A deep neural network-based smart error measurement method for fiscal accounting data

The error measurement of fiscal accounting data can effectively slow down the change of financial assets. Based on deep neural network theory, we constructed an error measurement model for fiscal and tax accounting data, and we analyzed the relevant theories of fiscal and tax performance evaluation. By establishing a batch evaluation index of finance and tax accounting, the model can monitor the changing trend of the error of urban finance and tax benchmark data scientifically and accurately, as well as solve the problem of high cost and delay in predicting the error of finance and tax benchmark data. In the simulation process, based on the panel data of credit unions, the entropy method and a deep neural network were used to measure the fiscal and tax performance of regional credit unions. In the example application, the model, combined with MATLAB programming, calculated the contribution rate of regional higher fiscal and tax accounting input to economic growth. The data show that the contribution rates of some fiscal and tax accounting input, commodity and service expenditure, other capital expenditure and capital construction expenditure to regional economic growth are 0.0060, 0.0924, 0.1696 and -0.0822, respectively. The results show that the proposed method can effectively map the relationships between variables.