A machine learning enhanced time series model for Nigerian unemployment rate prediction
DOI:
https://doi.org/10.64497/jssci.79Keywords:
unemployment rate, ARIMA, XGBoost, hybrid model, time series forecastingAbstract
Reliable unemployment rate forecasting is essential for effective economic planning and labor market interventions, particularly in developing economies like Nigeria. This study proposes a hybrid modeling approach that integrates the Autoregressive Integrated Moving Average (ARIMA) model and the Extreme Gradient Boosting (XGBoost) algorithm within a stacked ensemble framework. Annual unemployment data from 1972 to 2014 were used. Model performance was evaluated via Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and coefficient of determination (R²). The results show that the hybrid ARIMA–XGBoost model significantly outperforms individual models, achieving RMSE = 1.1767, MAE = 0.9602, MAPE = 0.0567, and R² = 0.9648. The findings demonstrate the effectiveness of combining linear and nonlinear modeling approaches in data-constrained environments.
Downloads
References
[1] A. Okoh and A. A. Akintunde, “Oil price volatility and unemployment in Nigeria: A nonlinear approach,” CBN Journal of Applied Statistics, vol. 13, no. 2, pp. 69–93, 2022, doi: 10.33429/cjas.13222.4/8
[2] M. O. Adenomon, “Modelling and forecasting unemployment rates in Nigeria using ARIMA model,” FUW Trends in Science & Technology Journal, vol. 2, no. 1B, pp. 525–531, 2017.
[3] I. A. Ogunsola, O. M. Olayiwola, and B. O. Sowola, “Modelling unemployment rates in Nigeria,” Asian Journal of Mathematical Sciences, vol. 4, no. 2, pp. 4–9, 2020.
[4] C. Akobi and A. Ogunmola, “Application of ARIMA methods on unemployment and inflation rates in Nigeria,” Asian Journal of Science and Technology, vol. 2, no. 6, pp. 843–869, 2024, doi: 10.58578/AJSTEA.v2i6.4087. DOI: https://doi.org/10.58578/ajstea.v2i6.4087
[5] J. O. Odhiambo, P. Weke, and P. Ngare, “A hybrid ARIMA-XGBoost model for time series forecasting with improved accuracy,” Asian Journal of Probability and Statistics, vol. 26, no. 4, pp. 27–44, 2024, doi: 10.9734/ajpas/2024/v26i4597. DOI: https://doi.org/10.9734/ajpas/2024/v26i10662
[6] P. Li and J.-S. Zhang, “A new hybrid method for China’s energy supply security forecasting based on ARIMA and XGBoost,” Energies, vol. 11, no. 7, p. 1687, 2018, doi: 10.3390/en11071687. DOI: https://doi.org/10.3390/en11071687
[7] M. Güler, A. Kabakçı, Ö. Koç, E. Eraslan, K. H. Derin, R. Ünlü, Y. S. Türkan, and E. Namlı, “Forecasting of the unemployment rate in Turkey: Comparison of the machine learning models,” Sustainability, vol. 16, no. 15, p. 6509, 2024, doi: 10.3390/su16156509. DOI: https://doi.org/10.3390/su16156509
[8] T. W. Ho, “Forecasting unemployment via machine learning: The use of average windows forecasts,” SSRN Electronic Journal, 2021. [Online]. Available: https://ssrn.com/abstract=3496138
[9] O. J. Olawale and D. J. Adashu, “A combination of ARIMA models and neural networks in forecasting Nigerian exchange rate,” African Multidisciplinary Journal of Sciences and Artificial Intelligence, vol. 1, no. 1, pp. 1–13, 2024, doi: 10.55810/2313-0083.1001. DOI: https://doi.org/10.58578/amjsai.v1i1.3367
[10] F. J. Ibrahim, I. S. Ahmad, H. A. Umar, N. B. Rabiu, A. S. Bichi, and A. M. Ahmad, “Prediction of unemployment rates with time series and machine learning techniques,” in Proceedings of the 2023 International Conference, Ahmadu Bello University, Zaria, Nigeria, 2023, pp. 183–187.
[11] World Bank, The Exposure of Workers to Artificial Intelligence in Low-Income Countries. Washington, DC, USA: World Bank, 2023. [Online]. Available: https://openknowledge.worldbank.org/entities/publication/4a11a37d-149a-44fb-a941-100065ff5eb8
[12] G. E. P. Box and G. M. Jenkins, Time Series Analysis: Forecasting and Control. San Francisco, CA, USA: Holden-Day, 1970.
[13] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, San Francisco, CA, USA, 2016, pp. 785–794. DOI: https://doi.org/10.1145/2939672.2939785
[14] T. Chai and R. R. Draxler, “Root mean square error (RMSE) or mean absolute error (MAE)?—Arguments against avoiding RMSE in the literature,” Geoscientific Model Development, vol. 7, no. 3, pp. 1247–1250, 2014, doi: 10.5194/gmd-7-1247-2014. DOI: https://doi.org/10.5194/gmd-7-1247-2014
[15] C. J. Willmott and S. M. Robeson, “Climatology and the evaluation of climate models,” International Journal of Climatology, vol. 25, no. 5, pp. 693–712, 2005, doi: 10.1002/joc.1181. DOI: https://doi.org/10.1002/joc.1181
[16] A. de Myttenaere, B. Golden, B. Le Grand, and F. Rossi, “Mean absolute percentage error for regression models,” Neurocomputing, vol. 192, pp. 38–48, 2016, doi: 10.1016/j.neucom.2015.12.114. DOI: https://doi.org/10.1016/j.neucom.2015.12.114
[17] N. J. D. Nagelkerke, “A note on a general definition of the coefficient of determination,” Biometrika, vol. 78, no. 3, pp. 691–692, 1991, doi: 10.1093/biomet/78.3.691. DOI: https://doi.org/10.1093/biomet/78.3.691
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Babafemi D. Ogunbona, Folorunsho O. Balogun , Samuel O. Olagunju
This work is licensed under a Creative Commons Attribution 4.0 International License.
- Abstract 785
- PDF 178
