Enhancing Decision-Making through Machine Learning and eXplainable AI: Case Studies in Credit Risk and Education

Advances in data collection and processing technologies, along with the rapid growth of computational power, have driven the development of numerous analytical techniques, often made transparent and understandable with explainable AI (XAI) and data explainability in general. However, the adoption of these techniques across different business domains remains challenging, as each domain presents unique characteristics that require careful consideration. This dissertation aims to address these challenges by demonstrating how data, AI, explainability and their subsets (anomaly detection, interactive machine learning, and XAI) can yield novel insights and enhance decision-making in areas such as credit risk management and education.