EXPLAINABLE AI IN DATA SCIENCE: BRIDGING TRANSPARENCY AND PERFORMANCE

Abstract

The rise of Artificial Intelligence (AI) has transformed data science, enabling unprecedented levels of prediction and automation. However, many high-performance AI models operate as "black boxes," lacking interpretability—a critical concern in sectors like healthcare, finance, and law. Explainable AI (XAI) aims to bridge this gap by enhancing model transparency while retaining predictive power. This paper explores the foundations, methodologies, and applications of XAI in data science. It highlights the balance between model accuracy and human understanding and identifies future directions for ethical and interpretable AI integration in decision systems.