Reinforcement learning-based sustainable educational intervention: An intelligent decision-making paradigm for global poverty reduction
Mingxuan Du, Lin Wang, Mario Di Nardo · 2025
Globally, there are many problems related to poverty and educational inequality. Because they are static and cannot easily adjust to changing situations, traditional approaches to better education frequently fail. In order to solve this issue, this paper uses reinforcement learning to create a sustainable educational intervention model. The intention is to use reinforcement learning in order to intelligently distribute educational resources and assist lessen poverty on a global scale by making better decisions. It is a model that the research develops to demonstrate how poverty and education are related. The model takes into account a number of variables, including socioeconomic status, poverty rates, and educational attainment. Then, in order to maximize the distribution of resources, a multi-agent decision-making method is created, in which families, governments, and schools collaborate. The study then presents sustainability indicators, such as environmental, and economic factors. These indicators are included in the reinforcement learning reward system, making sure that policies are always effective in the long run and able to adapt to the surroundings. In the experimental part, the model is tested by means of actual data and simulations, contrasting with more recent approaches and conventional approaches. The performance of the various approaches by using the MADDPG performs significantly better than the others. It particularly increases the amount of money spent by the government and has more measurable effects on the environment, particularly with regard to the use of resources and the efficacy of the entire policy. This work makes a useful and sensible approach to combating poverty on a global scale by moving from fixed to flexible solutions with the use of instructional tactics. Additionally, it offers helpful technical advice to support the Sustainable Development Goals (SDGs) of the UN.