Abstract
In this study, the quantity of municipal solid waste (MSW) generated in Istanbul between 1996 and 2008 was modeled. Statistical models such as linear regression, nonlinear regression, and time series and radial basis network, one of artificial intelligence techniques, were used. In the training period, average error percentages of four different models were at least 6.03% and at most 8.73%. After the training period, these four models were used to predict quantities of MSW in Istanbul which will be generated between 2009 and 2020. The model named random walk with drift (with 6.03% average error) was proposed as the best appropriate model predicting quantities of MSW in Istanbul. By increasing the number of model inputs, developing such a dynamic model predicting quantities of MSW in Istanbul and dissemination of this developed model at national scale are in the future goals of this study.