Artificial neural networks for assessing waste generation factors and forecasting waste generation: A case study of Chile

One of the bottlenecks in implementing waste management policies in Chile is the lack of information on factors correlating with waste generation. Recognising these factors is essential for implementing policies to reduce waste generation. From over 40 global variables indicating demographic, socio-economic and climatic conditions, Population, Percentage of Urban Population, Years of Education, Number of Libraries, and Number of Indigents were identified as the most important factors correlating with waste generation in Chile, all relating positively. Using these variables, communes were clustered into groups from which representative communes were selected for further data collection for forecasting waste generation at a communal level. Artificial Neural Networks were used for identifying factors, clustering communes and forecasting waste generation. The model is designed to represent most of the communes of a country. In this study, the best scenario represents 67.3% of the communes, based on the representativeness of each selected representative. However, due to lack of information, this rate decreased to 48.8%. Forecasted rates show that by 2010, representative communes will generate 100, 240 and 2,900 tonnes/month, with yearly variation rates of less than 1%. These predictions will be used to obtain estimates for each represented group and a significant portion of Chile.