Random Utility

Random utility models (RUMs) are a cornerstone of discrete choice modeling, aiming to understand and predict how individuals choose among multiple options based on their perceived utilities, which inherently include random components. Current research focuses on improving RUMs' flexibility and interpretability, exploring novel architectures like neural networks and gradient boosting to capture complex, non-linear utility functions while maintaining economic consistency and incorporating domain knowledge. These advancements enhance the accuracy and applicability of RUMs across diverse fields, from recommendation systems and transportation planning to social choice theory and behavioral economics, by providing more realistic and insightful models of decision-making.