Abstract
Accounting for finances is maintaining records of analyzing financial activities to produce financial declarations, income, and counting cash flow, and balance sheets. Verifying the completeness and correctness of financial accounts is the procedure known as auditing. Auditors guarantee the accuracy of financial data supplied by accounting and the integrity of accounting procedures. This study determines to establish an artificial intelligence model, named Dynamic Sea Lion Optimization (DSLO), for enhancing the efficiency of financial accounting and auditing processes. In this study, taxpayers functioning under an average tax regime, efficiency, and grow a fraud forecast method based on Dynamic Sea Lion Optimized Efficient Random Forest (DSLO-ERF). A comprehensive dataset is used to evaluate predictive models for detecting tax fraud. The data was preprocessed using normalization for the obtained data. Fraud, audit, administrative cost sharing, and external economic activity among the significant fraud predictors. The proposed framework is tested from 2018 to 2022 on the whole population of taxpayers and profit. In a comparative analysis, the proposed method performs various evaluation metrics. The result revealed that compared with other current methods, the efficacy of the suggested technique is better in terms of accuracy (98.621%), precision (99.517%), and recall (99.612%) in this condition, the proposed DSLO-ERF method also had the quickest training time of 0.10 seconds. The research focuses on an innovative approach to achieving greater operational efficiency and financial integrity, as well as the strategic advantages of AI adoption in the financial industry.
Keywords
Get full access to this article
View all access options for this article.
