ANALYZING PLASTIC WASTES IN AFRICA WITH MACHINE LEARNING

Plastic pollution is a global environmental crisis, and Africa is no exception.As the continent undergoes rapid development, plastic consumption has soared, resulting in a surge in plastic waste.To effectively address this pressing issue, understanding and modelling plastic pollution in Africa is essential.This thesis investigated plastic modelling in the African context, examining the intricacies of the problem, the methodologies employed, and the potential solutions to mitigate the impact of plastic waste on the continent.From the global dataset of plastic waste of the world from 1950 to 2010, we carved out African waste dataset and analysed with machine learning techniques.Support Vector machines, catboat algorithm, logistic regression algorithm are the foremost Algorithms used.To provide a comprehensive review of sources and characteristics of plastic wastes typically found in Africa.Identified a plastic classification method for easier plastic identification and waste modelling called region of plastic waste mismanagement.A review of current industry practices and research regarding plastic waste modelling was also explored to provide an outline for a conceptual framework for complete plastic waste management.This study's findings will be useful to industry practitioners and related software vendors in establishing improved practices and tools for plastic management and look-ahead scheduling.This method shows an accuracy of 97.8% with SVM, 94.2%, with catboost while other algorithms like firefly and logistic regression showed 73.6% and 65.3%, respectively.