Panel Dataset on Conflict, Illicit Practices, and Socio-Economic Development

This dataset contains a panel of indicators used for modelling and forecasting socio-economic vulnerabilities in conflict-affected countries using GRU-based neural networks. It integrates three groups of variables: conflict impact indicators (e.g., displaced persons and refugees), illicit practices indicators (corruption, AML risks, and cyber vulnerabilities), and socio-economic development indicators (macroeconomic, demographic, labour market, and stability measures). The dataset covers 135 countries for the period 2012–2024 and was constructed after data cleaning and consistency checks to ensure reliability for machine learning applications.