Predicting waste management system performance from city and country attributes

Supporting good waste management practices is crucial for the sustainable development of cities. Transforming the practices of cities is a complex problem that requires understanding their societal, technological, and economic processes. Machine learning is a branch of artificial intelligence techniques that generates models from patterns in data. Many of these models are difficult to present to city planners because of poor transparency. To provide insights for policymaking, interpretable machine learning models between city attributes and waste management performance are needed. Country attributes have a top-down influence on the sustainability of cities. Their inclusion provides deeper insights in addition to city-wide scope analysis. This work develops a rule-based machine learning model in the impact of city and country attributes on waste management. Rough setbased machine learning is used to generate models consisting of if-then rules with data from 100 cities in 41 countries. The results identify local governance, employment, and technological research as core attributes that influence sustainable waste management. The rough set-based machine learning models attained binary classification accuracies of 89%-91%. The implications on waste management and Circular Economy transition policies are discussed in this study.

Automated summary

Supporting good waste management practices is crucial for the sustainable development of cities. Machine learning models can provide insights for policymaking by analyzing the relationship between city attributes and waste management performance. This study develops rough set-based machine learning models and identifies local governance, employment, and technological research as core attributes that influence sustainable waste management.