EXPLAINABLE ARTIFICIAL INTELLIGENCE

Catalog description

This course introduces the principles and methods for designing interpretable, transparent, and trustworthy AI systems. Students study both model-agnostic and model-specific explainability techniques for machine learning and deep learning, including feature attribution, saliency mapping, local surrogate models, counterfactual reasoning, and post-hoc interpretation. Emphasis is placed on the trade-offs between accuracy, transparency, fairness, and accountability in real-world domains such as healthcare, finance, and autonomous systems. Through hands-on projects and critical analysis, students learn to design and evaluate explainable AI models that support responsible decision-making. (3 credits) fall, spring, summer