Barigou, KarimBaltes, MartinMartinBaltes2025-07-082025-07-082025-05-2720252025-05-27https://hdl.handle.net/2078.2/43535Flooding poses a significant risk to businesses, leading to temporary closures, reduced productivity, and financial losses. This study leverages predictive modeling techniques to estimate the number of business interruption days following flood events. Using an anonymous dataset of insurance claims, various methodologies, including Generalized Linear Models (GLM), Generalized Additive Models (GAM), Classification and Regression Trees (CART), Random Forest (RF), and Gradient Boosting Machines (GBM) are assessed for their effectiveness in analyzing climate-induced disruptions. The research applies these models to a case study, examining the flood vulnerability of fuel stations, refineries, and polymerization centers across Belgium. Through an extensive review of meteorological data, historical precipitation trends, and terrain-specific flood risks, the study estimates potential financial losses due to business interruptions. By combining data-driven approaches with real-world applications, this work contributes to improving climate risk assessment, enhancing disaster preparedness strategies, and understanding broader economic consequences of extreme weather events.text::thesis::master thesis