Global Plastic Waste Management: Analyzing Trends, Economic and Social Implications, and Predictive Modeling Using Artificial Intelligence

Plastic waste, which is a result of human activities in America, has become one of the most critical environmental issues in the 21st century, and this calls for an urgent prescription of strategies at a global management level. The pervasiveness of plastic in modern life has created an unparalleled surge in plastic waste, which, unless adequately managed, is poised to pose severe threats to ecosystems, human health, and the global economy. The utmost objective of this study was to perform an extensive analysis of global plastic waste management practices in the USA, with a specific concentration on pinpointing the economic and social implications of these practices. This research project therefore intends to probe into the waste management practice applied in different countries for understanding the various best practices, challenges, and areas of improvement. The research project also aimed to employ AI-driven predictive models, notably, gradient boosting algorithm, linear regression, and random forest to predict the future trends in plastic wastes generated. Diverse datasets were used, to ensure that the study of global plastic waste management practices was comprehensive. Primary data on the conditions of global plastic waste generation was obtained through the World Bank's database, which provides detailed data on waste composition, generation rate, and methods of disposal in many countries. Also, the sources of economic indicators were OECD reports and UNEP publications on the hidden economic costs of plastic waste to municipal budgets. Data on its social impact, such as health effects and metrics involving environmental pollution, were provided by the World Health Organization through studies it conducted along with reports from environmental NGOs such as Greenpeace. The Gradient Boosting model performed the best with relatively high accuracy, followed by Logistic Regression and Random Forest Classifier. Besides, the Gradient Boosting model offered the highest Macro Average F1-score, which suggests better overall performance in balancing precision and recall for all classes. Predictive insights provided by the proposed models are valuable tools to expect future trends and patterns in plastic waste generation. Advanced analytics and machine learning can help predict the volume of plastic waste generated across different geographies and sectors. Application of the predictive models in plastic waste management contexts has huge potential about information and shaping of policy decisions. Predictive analytics can use historical trends on production, consumption, and generation of waste and recycling rates of plastics to create forecasts about the future and define high-risk areas.