Paper ID: 1233

A Comparative Study of Advanced Machine Learning and Deep Learning Models for Municipal Solid Waste Forecasting: A Case Study of Surat, India 

 

Enlai Zhang, Jiaqiang Jiang*, & Zijuan Ren

Abhijit R. Rathod1, *, & Vinodkumar M. Patel2

1Gujarat Technological University, Ahmedabad, Gujarat, India 

2Shantilal Shah Engineering College, Bhavnagar, Gujarat, India

 

*Corresponding author: arrathod1709@gmail.com

 

Abstract

To make the city planning sustainable, especially in rapidly growing cities of the world like Surat in India, the implementation of effective waste management is crucial.  The primary factor governing this is the ability to accurately predict the quantity of municipal solid waste (MSW) likely to be generated. This study presents a comprehensive comparative analysis of various predictive models including linear regressions, kernel approaches, gradient boosting as well as the deep learning architectures. Using historical data from Surat, the study applies rigorous feature engineering across more than 400 features related to temporal data, socio-economic conditions, climatic variables, COVID-19 impact, and mobility patterns. All these features were processed through systematic pre-processing, data cleaning, and feature engineering. The novel contribution of this study is the systematic feature engineering framework that explicitly encodes temporal structure (419 engineered features including lagged values, rolling statistics, and seasonal decomposition), enabling simple linear models to capture complex waste generation patterns. Ten distinct models, ranging from statistical approaches to machine learning and deep learning were evaluated and compared. Advanced ensemble models, including LightGBM (R² = 0.983), CatBoost (R² = 0.977), and XGBoost (R² = 0.970) demonstrated strong performance. The best-performing models (OLS and Gaussian Process Regression) achieved R² = 0.997 with Mean Absolute Percentage Error (MAPE) = 1.43%. In this study, linear models trained within a few milliseconds and achieved per-sample inference times on the order of 0.004-0.008 ms, whereas the tuned MLP and tree-based ensembles required seconds of training and millisecond-level inference, corresponding to differences of roughly two to three orders of magnitude in computational cost. Other notable performers include Lasso regression (R² = 0.979), tuned MLP (R² = 0.968), and Random Forest (R² = 0.960). All the results clearly demonstrate that the quality of feature engineering is more impactful than the model complexity. Here, simple OLS with engineered features (R2=0.997) clearly outperformed deep learning architectures (MLP R2=0.966) by approximately 3.1%, while also achieving faster prediction speed. The detailed analysis of feature relevance shows the importance of features like lagged MSW values, rolling statistics, different demographic metrics, event-driven factors like sudden changes observed in festival times, impact of the COVID 19 lockdown. The research finds that adoption of systematically designed feature engineering framework is a valuable tool for MSW management. This study enhances the developing field of MSW forecasting by providing a thorough benchmarking of models and presents the practical recommendations for the policymakers aiming at sustainable growth of the city.

Keywords: deep learning; forecasting; gradient boosting; machine learning; municipal solid waste; Surat; time series.

References

Abbasi,  M., & El Hanandeh, A. (2016). Forecasting municipal solid waste generation using artificial intelligence modelling approaches. Waste Management, 56, 13–22. https://doi.org/10.1016/j.wasman.2016.05.018

Abhishek. (2020; October 12). Imputing a missing value with a constant for a categorical data. Retrieved from Stack Overflow website: https://stackoverflow.com/questions/64308656/imputing-a-missing-value-with-a-constant-for-a-categorical-data

Aleksey, B. (2018). Simple techniques for missing data imputation. Retrieved July 16, 2025, from https://kaggle.com/code/residentmario/simple-techniques-for-missing-data-imputation

Bergmeir, C., Hyndman, R. J., & Benítez, J. M. (2016). Bagging exponential smoothing methods using STL decomposition and Box–Cox transformation. International Journal of Forecasting, 32(2), 303–312. https://doi.org/10.1016/j.ijforecast.2015.07.002

Billal, M. M., & Kumar, A. (2025). Forecasting residential and nonresidential solid waste generation, disposal, and diversion using three machine learning approaches. Biofuels; Bioproducts and Biorefining. https://doi.org/10.1002/bbb.70010

Breiman, L. (2001). Random Forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324

Cerqueira, V., Moniz, N., & Soares, C. (2020). VEST: Automatic Feature Engineering for Forecasting (Version 1). Version 1. arXiv. https://doi.org/10.48550/ARXIV.2010.07137

Chatterjee, D. U. (2024). Urbanization and Solid Waste Management: It’s Impacts on Human Health and Environment in Asansol Town. Bioscene.

Chen, J., Cui; Y., Wei, C., Polat, K., & Alenezi, F. (2025). Advances in EEG-based emotion recognition: Challenges; methodologies; and future directions. Applied Soft Computing, 180, 113478. https://doi.org/10.1016/j.asoc.2025.113478

Chen, J., Cui; Y., Wei, C., Polat, K., & Alenezi, F. (2026). Driver fatigue detection using EEG-based graph attention convolutional neural networks: An end-to-end learning approach with mutual information-driven connectivity. Applied Soft Computing, 186, 114097. https://doi.org/10.1016/j.asoc.2025.114097

Chen, J., Fan, F.,Wei, C., Polat, K., & Alenezi, F. (2025). Decoding driving states based on normalized mutual information features and hyperparameter self-optimized Gaussian kernel-based radial basis function extreme learning machine. Chaos, Solitons & Fractals, 199, 116751. https://doi.org/10.1016/j.chaos.2025.116751

Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; 785–794. New York; NY; USA: Association for Computing Machinery. https://doi.org/10.1145/2939672.2939785

Chen, Y.-C. (2018). Effects of urbanization on municipal solid waste composition. Waste Management, 79, 828–836. https://doi.org/10.1016/j.wasman.2018.04.017

Duong, T. (2024; August 14). Introduction to Feature Engineering for Time Series Forecasting. Retrieved July 13, 2025, from https://www.linkedin.com/pulse/introduction-feature-engineering-time-series-duong-truong-kinic

Feature‑engine developers. (2025, January 22). Feature-engine—1.8.3. Retrieved July 16, 2025; from https://feature-engine.trainindata.com/en/latest/

GeeksforGeeks. (11:43:45+00:00). One Hot Encoding in Machine Learning. Retrieved July 16, 2025; from GeeksforGeeks website: https://www.geeksforgeeks.org/machine-learning/ml-one-hot-encoding/

Gordon, J. (2023, June 20). Practical Guide for Feature Engineering of Time Series Data. Retrieved July 13, 2025; from dotData website: https://dotdata.com/blog/practical-guide-for-feature-engineering-of-time-series-data/

Kamalov, F., & Sulieman, H. (2021, October 25). Time series signal recovery methods: Comparative study. arXiv. https://doi.org/10.48550/arXiv.2110.12631

Lewinson, E. (2022, February 17). Three Approaches to Encoding Time Information as Features for ML Models. Retrieved August 30; 2025; from NVIDIA Technical Blog website: https://developer.nvidia.com/blog/three-approaches-to-encoding-time-information-as-features-for-ml-models/

Maximilian, C. (2025, February 16). Rolling/Time series forecasting. Retrieved July 13, 2025, from Https://tsfresh.readthedocs.io/en/latest/index.html website: https://tsfresh.readthedocs.io/en/latest/text/forecasting.html

Pasa, L., Angelini, G.,  Ballarin, M., Fedrizzi; P., & Sperduti, A. (2025). Enhancing door-to-door waste collection forecasting through ML. Waste Management, 194, 36–44. https://doi.org/10.1016/j.wasman.2024.12.044

ProjectPro. (2023; April 12). How to Impute Missing Values with Mean in Python? -. Retrieved July 16, 2025, from ProjectPro website: https://www.projectpro.io/recipes/impute-missing-values-with-means-in-python

Raschka, S., & Mirjalili, V. (2020). Python machine learning: Machine learning and deep learning with Python, scikit-learn, and TensorFlow 2 (3rd ed). Birmingham: Packt Publishing.

Sheskin, D. J. (2020). Handbook of Parametric and Nonparametric Statistical Procedures; Fifth Edition (5th ed.). New York: Chapman and Hall/CRC. https://doi.org/10.1201/9780429186196

Smyth, S. (2024). Waste Management in Developing Countries: Challenges and Solutions. Advances in Recycling & Waste Management, 9(03). https://doi.org/10.37421/2475-7675.2024.9.350

The MathWorks; Inc. (2025, July 16). Rolling-Window Analysis of Time-Series Models—MATLAB & Simulink. Retrieved July 13, 2025; from Rolling‑Window Estimation of State‑Space Models website: https://in.mathworks.com/help/econ/rolling-window-estimation-of-state-space-models.html

UN-Habitat. (2018). Indicator 11.6.1 Training Module: Solid Waste in Cities. Nairobi: UN Human Settlements Programme. Retrieved from UN Human Settlements Programme website: https://unhabitat.org/sites/default/files/2019/02/Indicator-11.6.1-Training-Module_Solid-waste-in-cities_23-03-2018.pdf

Vu, H. L., Ng, K. T. W., Richter; A., & Kabir, G. (2021). The use of a recurrent neural network model with separated time-series and lagged daily inputs for waste disposal rates modeling during COVID-19. Sustainable Cities and Society, 75, 103339. https://doi.org/10.1016/j.scs.2021.103339

Wei, C., Alenezi, F., Chen, J., Wang, H., & Polat, K. (2026). Nonlinear Feature Decomposition and Deep Temporal–Spatial Learning for Single-Channel sEMG-Based Lower Limb Motion Recognition. IEEE Sensors Journal, 26(3), 4120–4126. https://doi.org/10.1109/JSEN.2025.3644160

Winastwan, R. (2024; March 25). Machine Learning Forecasting of Time Series. Retrieved July 13, 2025, from Train in Data’s Blog website: https://www.blog.trainindata.com/machine-learning-forecasting/

World Bank Group. (2022). Solid Waste Management. World Bank Group. Retrieved from World Bank Group website: https://www.worldbank.org/en/topic/urbandevelopment/brief/solid-waste-management

Zhang; Z., Chen, Z., Zhang,  J., Liu, Y., Chen; L., Yang; M., … Yap, P.-S. (2024). Municipal solid waste management challenges in developing regions: A comprehensive review and future perspectives for Asia and Africa. Science of The Total Environment; 930, 172794. https://doi.org/10.1016/j.scitotenv.2024.172794