A hybrid AHP and ensemble machine learning based approach for optimising dumpsite selection in Raiganj municipality for sustainable urban waste management

The rapid urbanisation of Raiganj Municipality generates a large quantity of solid waste daily, which indicates the necessity of a scientifically robust approach to dumpsite location selection to mitigate environmental and public health risks. The present research combines the Analytic Hierarchy Process (AHP) with machine learning techniques, specifically Random Forest, Logistic Regression , and Support Vector Machine to analyse samples (80 spots), 21 variable binary datasets for the selection of suitable dumpsite locations. The dataset captures the critical spatial factors like distance from the water bodies, Distance from the present open dumpsite, slope, and land use and land cover, etc., to classify regions as environmentally low, moderate, and high Susceptibility regions. This framework enabled a Spatial Zoning and Site Regulation Suggestion allocating Low Susceptibility regions (Wards 25, 23, 24 and others = 2.55 sq. km) for legal dumping , except High Susceptibility regions (Wards 13, 14, 12, 16, 15 and others = 3.81 sq. km) around Kulik River for its Susceptible infrastructure and with the highest accuracy (0.89) and F1-score (0.86), the Random Forest model strengthened the ensemble classification. While in Moderate Susceptibility regions (Wards 1, 2, 10, 18 and others = 4.39 sq. km) where Conditional dump might be allowed. Although the model lacked information regarding site conditions, it still utilised accurate outputs to establish safe waste routes and enforce buffer distances, as well as includes community input to develop an environmentally compliant framework for selecting designated dumpsites throughout the city of Raiganj that is adaptable to any urban environment.