Yemeni Synthetic Society

A synthetic society that replicates the characteristics of the local society can provide a valuable testing ground for the purpose of building an agent-based simulation model aimed at assessing and extrapolating the impact of Qat consumption of Yemeni society regarding to health, economy, and social aspects, as well as predicting the effects of local public policies on Qat consumption in Yemen. While synthetic society datasets are available for several countries, there is currently no synthetic society data for Yemen. This study proposes the creation of a synthetic society dataset for individuals at a detailed level in Yemen. This synthetic society was generated using a genetic algorithm based on multiple census data, several general statistical studies, and surveys on households and Qat consumption in Yemen, which include detailed or aggregated socio-demographic attributes such as age, sex, income, education level, employment status, stage of consumption, etc. We designed an efficient fitness function computation based on a hybrid distance, enabling us to incorporate over one hundred statistics and generate an artificial society comprising more than one hundred thousand artificial agents with an SRMSE of 0.231 compared to the statistics of the real society.<br><br><br>