Research Information System
2nd International Conference on Forthcoming Networks and Sustainability in the IoT Era (FoNeS-IoT), ELECTR NETWORK, 8 - 09 January 2022, vol.129, pp.157-164, (Full Text)
The previous accounting information quality evaluation only used simple processing of accounting information indicators, such as averaging or artificially giving the weight of each indicator to weighted summation. The evaluation results are very subjective. Use BP network to establish a model of the accounting information quality evaluation system, obtain accounting information evaluation indicators through investigation and analysis, quantify them into definite data as its input, use BP neural network to evaluate the actual output, and use the previously obtained accounting information effect. As the desired output. When the error reaches the desired minimum value, the evaluation is considered successful. After the evaluation is successful, more accurate weights and thresholds can be obtained, and the network after the evaluation is successful is used to process another set of newly obtained accounting information evaluation indicators to obtain the accounting information quality evaluation results. This method is used in the evaluation of accounting information quality, which not only overcomes the subjective factors of experts in the evaluation process, but also obtains satisfactory evaluation results, which has a wide range of applicability.