An Integrated Model of Data Envelopment Analysis and Artificial Neural Networks for Improving Efficiency in the Municipal Solid Waste Management

Abstract

In the last decades, integral municipal solid waste management (IMSWM) has become one of the most challenging areas for local governmental authorities, which have struggled to lay down sustainable and financially stable policies for the sector. In this paper a model that evaluates the efficiency of IMSWMs through a combination of Data Envelopment Analysis (DEA) and an Artificial Neural Network (ANN) is presented. In a first stage, applying DEA, municipal administrations are classified according to the efficiency of their garbage processing systems. This is done in order to infer what modifications are necessary to make garbage handling more efficient. In a second stage, an ANN is used for predicting the necessary resources needed to make the waste processing system efficient. This methodology is applied on a toy model with 50 towns as well as on a real-world case of 21 cities. The results show the usefulness of the model for the evaluation of relative efficiency and for guiding the improvement of the system.