A Performance Evaluation of Machine Learning Algorithms for Fraud Detection in Accounting in SMEs

Small and medium-sized enterprises (SMEs) often struggle to detect accounting fraud, threatening their financial stability and reputations. This study conducts a comparative analysis of machine learning algorithms to identify the most effective models for detecting anomalies in accounts payable. We evaluated algorithms such as Random Forest, Decision Tree, and XGBoost, training them on historical financial data and assessing their performance using cross-validation techniques. Evaluation metrics included precision, recall, F1-score, and accuracy, enabling a detailed analysis of each model's ability to detect suspicious patterns. The results indicate that Random Forest achieved the highest performance, with an accuracy of 97%, positioning it as the most effective option for fraud detection. This analysis supports the development of advanced risk management tools, enhances transparency in accounting processes, and helps SMEs make secure, data-driven decisions—ultimately strengthening their resilience against accounting fraud.