Time Series Forecasting of Solid Waste Generation in Arusha City - Tanzania

Abstract

Statistical time series modeling is widely used in prediction and forecasting studies. This study intends to analyze, compare and select the best time series model for forecasting amount of solid waste generation for the next years in Arusha city - Tanzania among ARMA/ARIMA and Exponential Smoothing models. The past data used are monthly amount of solid waste collected by the city authorities from year 2008 to 2013. The result indicated that ARIMA (1, 1, 1) outperformed other potential models in terms of MAPE, MAD and RMSE measures and hence used to forecast the amount of the solid waste generation for the next years.

Keywords: ARIMA models, Exponential Smoothing models, time series, MAPE, MAD, RMSE