The AI and Machine Learning Advances Journal works towards becoming a leading journal for AI/ ML research findings. In this way, it performs a function of connecting academic, industrial, top machine learning algorithms and governmental researchers to exchange know-how and innovations that are shaping the development of intelligent systems at the present time.
Article Views: 32
Artificial Intelligence (AI) is used in strategic decision-making; however, it is not clear whether the use of explainable machine learning leads to more managerial judgment when prediction accuracy is high. This study analyzes the proposed ‘Managerial Blind Spot Effect’ by combining predictive modelling with behavioural analysis applied to the IBM Telco Customer Churn dataset and the AI Decision Dependence and Cognitive Caution dataset. Several models, including Explanatory Boosting Machine (EBM), XGBoost, Random Forest and Neural Network, were tested, and the explanation produced by the native model EBM was enhanced by SHAP-based explanation. The behavioural results revealed that both Trust in AI and Cognitive Caution significantly contributed to an increase in AI Dependence. EBM delivered the best predictive performance with intrinsic transparency. The results highlight the importance of using explainable AI prediction models in conjunction with managerial judgment in dashboards.
Explainable Artificial Intelligence, Strategic Decision-Making, Managerial Blind Spot Effect, AI Dependence, Customer Churn Prediction