Forecasting Municipal Solid Waste (MSW) generation in Klang, Selangor using Artificial Neural Network (ANN)

• Published in: Malaysian Journal of Medicine and Health Sciences
• Main Author: Mohamad N.A.J.; Yatim S.R.M.; Abdullah S.; Azmin M.T.; Alwi N.
• Format: Article
• Language: English
• Published: Universiti Putra Malaysia Press 2022
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85134387462&doi=10.47836%2fmjmhs18.8.21&partnerID=40&md5=e7224bd6bdfd129870e2151c0a341ca5

Introduction: Municipal solid waste (MSW) generation involves complex mechanisms. Prediction of future MSW amount can help the authority to comprehend and produce guidelines towards disposal system. In this study, we aimed to develop non-linear model for MSW prediction in Klang, Selangor on monthly basis. Methods: Data of MSW were acquired from Klang Municipal Council, composed of monthly MSW collection from the year 2017 to 2021. Non-linear autoregressive (NAR) models using artificial neural network (ANN) were developed using MATLAB software, and the best-fitted model for MSW prediction is assessed via coefficient of determination (R2). Results: We found a fluctuating trend of MSW generated between months and years. MSW has statistically significant difference on monthly basis (p < 0.05), with 17 273 kg produced in January, and interestingly MSW has no statistically significant difference (p > 0.05) across the years, the highest was 21 7310.4 kg in the year 2021. The best architecture for the NAR models based on neurons testing ranges from 1 to 15 is 1-1-1, 1-2-1, 1-3-1,1-4-1, 1-5-1, 1-7-1, 1-8-1, 1-9-1, 1-10-1, 1-12-1 and 1-14-1 with R2 > 0.90. Conclusion: The developed models can be used for MSW prediction for MSW management by other municipalities. © 2022 UPM Press. All rights reserved.