Research data and code of the work titled "Forecasting Municipal Solid Waste Disposal Rates Using SARIMAX model"
Description
Recently, various methodologies have been employed to develop forecasts for municipal solid waste generation rates (MSWDR). Notably, several of these methods are highly robust, including dynamic systems, convolutional neural networks, machine learning, and deep learning. However, when only a relatively small dataset is available, the efficiency of these methods significantly diminishes. Consequently, simpler methods such as time series analysis prove to be more suitable under these conditions, as models like the Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) efficiently capture the seasonal patterns and trends in variables like the MSWDR. This dataset embodies the time series derived from public databases for the case study in Colombia. The .csv files contain the foundational data required for implementing the forecasts. The .R file is the script used in RStudio to conduct the research. The outcomes produced by executing this script outline the methodological framework adopted to select, validate, and deploy the SARIMAX model to forecast MSWDR. Author details and affiliations are omitted for the purpose of a double-blind review. These will be updated once the review process is completed.
Files
Steps to reproduce
1. Place all files in the same folder. 2. Run MSW_SARIMAX.R (RStudio)