Hybrid Artificial Intelligence Framework for Financial, HR, and Operational Management

Abstract

The use of AI to help organizational decision-making has become popular; yet, the financial, HR, and operational tools are designed separately. The current work presents a Hybrid Artificial Intelligence Framework for Financial, HR and Operational Management which combines predictive models tailored for each field with risk fusion and cross-domain rule-based decision support. The HR module uses Logistic Regression, Random Forest and XGBoost to analyze employee attrition. At the same time, the financial module predicts financial distress of a company using similar classification methods. The operational module utilizes machine learning to predict demand based on historical retail transaction data. According to the experimental findings, Logistic Regression reached the ROC-AUC of 0.8273 in HR attrition analysis, whereas XGBoost obtained ROC-AUC of 0.9540 in financial distress prediction. The best performance in terms of R² was demonstrated by Random Forest in operational predictions, reaching 0.4439. The results from the three modules are normalized and combined using a weighted risk fusion model. After that, the rule-based decision support system evaluates the risk of cross-domain combinations and offers coordinated managerial solutions.