Learning from waste: Machine Learning for health risk prediction and computer vision-based sorting in Ghana

This research explores the application of machine learning (ML) to address the public health and environmental challenges posed by solid waste disposal in Ghana. It validates a community-perceived link between waste disposal and illness patterns using predictive modeling and proposes computer vision for waste sorting to improve management practices. The study highlights the potential of data-driven approaches to inform policy and operational strategies for waste management and public health.

• Machine learning, specifically a Random Forest classifier, can predict illness categories based on waste disposal practices and demographic data, validating a community-perceived link.
• The predictive model achieved an Area Under the Curve (AUC) of 0.82 for predicting illness categories, demonstrating its efficacy in identifying health risks associated with waste.
• Computer vision (YOLOv8) models can classify common household waste types (organic, plastic, paper, metal, glass, fabric) with high accuracy, indicating potential for automated sorting.
• The most accurate YOLOv8 model achieved a mean Average Precision (mAP) of 0.707 over 6 classes, suggesting operational viability for waste segregation.
• Inappropriate solid waste disposal is a significant public health and environmental concern globally, with substantial economic costs and avoidable deaths.

Reported by arXiv — Computers and Society · International. This briefing summarises the publication for executive use; the document itself is not reproduced here.