Integrating Decision-Making Quality with Decision Support Systems: A Machine Learning Approach Based on Behavioural Insights from Business Professionals

Abstract

ABSTRACT  

Decision-making quality is essential for organizational success and individual satisfaction. However, assessing and predicting decision quality has often stayed at a theoretical level and has not been integrated into practical tools like decision support systems (DSS). This study aims to close that gap by developing an approach based on machine learning that combines behavioural factors influencing decision-making with explainable artificial intelligence (XAI) techniques. Data were collected from business professionals across various industries, focusing on individual, environmental, and organizational factors that shape their decision satisfaction and regret. The dataset was balanced using synthetic sampling to ensure strong model performance. An XGBoost model predicted whether individuals would make the same decision again under current conditions. The model’s performance was validated with ROC curves, confusion matrices, and SHAP analysis to identify the key factors that influence the outcomes. The findings showed that variables such as perceived organizational fairness, external pressures, personal experience, and intuitive processes had a significant effect on decision quality. Including these behavioural insights into DSS frameworks can improve the accuracy and practical relevance of these systems. This study contributes to the link between behavioural science, machine learning, and decision support, providing new perspectives for managers and system developers.